(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Yaughan and Curriboo Plantations: Studies in Afro American Archaeology"

V38/x 

i Yaughan an 
Studies in A 




iboo Plantations: 
merican Archaeology 




Soi 
Project 



Received 
APR 301984 



DOCUMENTS 



YAUGHAN AND CURRIBOO PLANTATIONS: 
STUDIES IN AFRO-AMERICAN ARCHAEOLOGY 



Funding provided by the 
Charleston District of the U.S. Army Corps of Engineers 



Contract Administered by: 

National Park Service 
Southeast Regional Office 
Archeologlcal Services Branch 
Contract Number C-5950(79) 



April, 1983 



by 

Thomas R. Wheaton 
Amy Friedlander 

Patrick H. Garrow 



Prepared by: 

Soil Systems, Inc. 
Marietta, Georgia Received 



Project Number 476-91219 



APR 3 1984 



DOCUMENTS 

'.' ■ . JRAFliES 



TABLE OF CONTENTS 

Abstract i 

Foreword ii 

Acknowledgements iii 

List of Figures iv 

List of Tables viii 

I. INTRODUCTION 1 

II. ENVIRONMENTAL SETTING 18 

Introduction 18 

Topography 18 

Soils 21 

Climate 23 

Habitats 24 

Natural Resources . 25 

III. HISTORIC OVERVIEW 30 

Introduction 30 

Protohistoric 30 

Early European Settlement, 1521-1663 34 

Population of Colonial South Carolina 35 

The Economic Development of the Low Country 38 

St. Stephen's Parish 43 

South Carolina and Cotton Production 1800-1860 50 

Slavery in South Carolina 54 

Ownership of Slaves 62 

After the Civil War 67 

IV. TESTING PHASE 71 

Introduction ., . 71 

Testing Goals 72 

Sites Recommended for Clearance 75 

Sites Recommended for Mitigation 86 

Analysis of Sites 388K75 and 38BK76 104 

Summary and Recommendations 113 

V. CURRIBOO, YAUGHAN, AND THE CORDES FAMILY 116 

Introduction 116 

Procedures 117 

Documents 118 

Summary of Ownership 120 

Yaughan Plantation 120 

Curriboo Plantation 123 

Cordes Family 126 



VI. MITIGATION METHODS 89 

Introduction 89 

Archaeological Field Methods 89 

Hand Excavations 92 

Mechanical Stripping 93 

Laboratory Methods 95 

VII. DESCRIPTIONS OF STRUCTURES AND FEATURES 98 

Introduction 98 

Descriptions 104 

Site 38BK75 104 

Summary of Site 388K75 129 

Site 38BK76 129 

Summary of Site 388K76 158 

Site 38BK245 158 

Summary of Site 38BK245 191 

Slave Architecture 193 

Size and Form 203 

Summary 211 

VIII. ARTIFACT ANALYSIS 212 

Introduction 212 

Ceramics 219 

Bottle Glass 250 

Gun Parts 252 

Tobacco Pipe Group 254 

Activities Group 256 

IX. ARTIFACT PATTERNS 265 

Introduction 265 

The Carolina and Frontier Artifact Patterns 266 

The Carolina Slave Artifact Pattern 277 

X. SLAVE SUBSISTENCE 287 

Introduction 287 

Ethnobotanical Remains 287 

Zooarchaeological Remains 293 

Conclusions 296 

XI. PLANTATION LIFE 298 

Introduction 298 

Plantation Life and Culture 299 

White Occupancy 299 

Economics of the Plantation 303 

The Impact of Inheritance and Sale 309 

The Black Community 318 

Conclusions 323 



XII. ACCULTURATION AT CURRIBOO AND YAUGHAN 325 

Introduction 325 

Defining the Problem 326 

Archaeology and History at Curriboo and Yaughan 327 

Criteria for Acculturation 329 

Acculturation Periods 330 

Arti factual Evidence 332 

Architectural Evidence 338 

Food Preparation 339 

XIII. CONCLUSIONS AND FUTURE RESEARCH QUESTIONS 342 

REFERENCES CITED 347 



APPENDIX A 
APPENDIX B 
APPENDIX C 
APPENDIX D 
APPENDIX E 
APPENDIX F 
APPENDIX G 
APPENDIX H 



Prehistoric Overview A-l 

Ownership of the Plantations 8-1 

Soil Tests C-l 

Architectural Historian's Report D-l 

Artifact Appendix . . . E-l 

Bead Analysis Report F-l 

Ethnobotanical Report G-l 

Faunal Report H-l 



ABSTRACT 



This report details an archaeological testing and mitigation effort conducted 
by Soil Systems Inc. at six historic sites in Berkeley County, South 
Carolina. The purpose of the project was to assess the significance of the 
sites and to mitigate the construction impacts of the U.S. Army Corps of 
Engineers' Cooper River Redi version Canal Project. Five sites were tested 
during spring 1979. Three of these sites and an additional site recommended 
by Interagency Archeological Services-Atlanta were subjected to intensive 
data recovery during the summer and fall of 1979. Twenty-nine structures and 
many associated features were examined at two eighteenth century plantations. 
Historical research produced evidence sufficient to characterize their slave 
occupations. The results of this research included extensive archaeological 
and historical data shedding light on eighteenth century African slaves. 
The conclusions offer new insights into the institution of slavery and offer 
a basis for future research. 



11 



FOREWORD 



"History," notes Ira Berlin in his preface to Slaves Without Masters , "is the 
study of exceptions." When it comes to the early history of Afro- Americans 
in this country the historical record is exceptional in its silence. This is 
one of the main reasons why historical archeology research in the seventies 
assumed as a mandated imperative the reconstruction of the Afro-American expe- 
rience, lifeways, culture and history from colonial times to the early 
twentieth century. 

The New York Times, in a Sunday editorial in the fall of 1982, touted arche- 
ology as "man's greatest library," and increasingly, governments are becoming 
aware of the potential of archeology to serve as the mediator between 
science, culture, and identity. It is in this context that we present this 
study of eighteenth century African slaves on the South Carolina coastal 
plain. The study is significant in a number of ways: in the differences it 
highlights between colonial and antebellum plantation systems; in its 
approach to the study of the process of acculturation as revealed in the 
archeological record; and finally in the contribution it makes to Black 
history by describing the early life of Afro- Americans as depicted in their 
own material remains. As James Deetz stated in summarizing the work at 
Parting Ways, "Since the arti factual and architectural remains of these 
communities are a better index of the life of Afro-Americans in their own 
terms, they hold great promise of supplementing American Black history in a 
different and important way... The archeology tells us that in spite of their 
lowly station in life they were the bearers of a life-style, distinctively 
their own, neither recognized nor understood by their chroniclers." 

We are proud to present this volume as the second in our series of profes- 
sional papers and we commend Pat Garrow, Tom Wheaton, Amy Friedlander and 
their colleagues at Soil Systems, Inc. for a job well done. The study 
represents a new reference standard in Afro-American archeology. 

We would also like to take this opportunity to applaud the commitment and 
financial support which the Charleston District, Corps of Engineers has given 
to these investigations. This volume, as well as the preceding monograph on 
The Mattassee Lake Sites , was undertaken as part of the cultural resources 
mitigation program on the Cooper River Rediversion Canal, and it is to the 
credit of the District, and the South Atlantic Division personnel that we 
were given the latitude to carry out the research in such an expansive and 
stimulating way. 



Victor A. Carbone 

Chief, Archeological Services, 

National Park Service 



m 



ACKNOWLEDGEMENTS 



The authors would like to take this opportunity to express their gratitude 
and to recognize the following individuals without whom the project could not 
have been completed. 

The permanent field crew consisted at various times of Christine Johnson, 
Assistant Field Director; David Babson, Field Lab Director; Linda France, 
John Green, Hunter Lesser, Jeanne Metropole, Lynda Morgan, Jenalee Muse, 
Aloyce Stewart, Charles Traylor, and Richard Warner. This crew was supple- 
mented from time to time by Jana Kellar, Howard Markel , Susan Markel , Steve 
Savage, Kim Savage, Stan Solomillo, Michael Watson, Claudia Watson, Ishmael 
Williams, and Melanie Younger. Some of the historic material was gathered by 
Linda Hart, Claudia Watson, and Ishmael Williams. 

We were also assisted in the field by James Thomson and Marlesa Gray of 
IAS-Atlanta; Marc Rucker, Louis Iacona, and Captain Bell of the Corps of 
Engineers, and Bill Dukes who helped with heavy equipment. 

The following individuals offered encouragement and advice which was greatly 
appreciated: Robert Stephenson, South Carolina State Archaeologist; Donald 
Sutherland, at that time archaeologist with the South Carolina State Historic 
Preservation Office; Paul Brockington, at that time with the Institute of 
Archeology and Anthropology; and Lei and Ferguson of the University of South 
Carolina. In the analysis phase we also received help and suggestions from 
Trlsh Logan then of the U.S. Forest Service, Helen Haskell of IAA, and Elaine 
Herold and Alan Lise of the Charleston Museum. 

The analysis phase was conducted by SSI personnel in Marietta who included 
Maria Almodovar, Linda France, and the assistant lab director, Lynda Morgan. 
This core staff was assisted by Pat BartHs, Ruth Caproni, Beth Gantt, Diane 
Garrow, and Margo Sellman. Drafting of maps and drawings was completed by 
Stephanie Low, Robert Robinson, Walter Rudolph, and Vincent Macek. We are 
also Indebted to Nancy Bechler for her valiant editorial effort. 

We would also like to express our thanks to all those from California to 
Scotland who courteously answered our letters of inquiry. We regret that 
space does not permit us to name each one separately. 

Individual research was conducted on four categories of remains. We wish to 
thank Elizabeth Reitz and Kay Wood of the University of Georgia for analyzing 
the faunal material; Paul Gardner of the University of North Carolina for his 
analysis of the floral material; Marvin Smith of the University of Georgia 
for his analysis of the glass beads; and Lane Greene of Greene Associates for 
research on mudwalled structures. 

Last, but certainly not least, we would like to express our appreciation to 
the people of St. Stephen who helped with logistics and information regarding 
the town's past and its environs. 



LIST OF FIGURES 

1. Project Area Location Map 2 

2. Historic Sites Location Map 11 

3. Parishes of South Carolina 1770 25 

4. Mouzon Map 1771 26 

5. Site 38BK91 Testing Site Plan 45 

6. Site 388K91 Test Unit 1 Profiles 46 

7. Site 38BK88 Site Plan 48 

8. Site 388K88 Test Units 1 and 5 50 

9. Site 38BK73 Site Plan 52 

10. Site 38BK73 Test Units 6 and 7 54 

11. Photos of Site 38BK73 and 388K75 55 

12. Site 38BK73 Block Excavation Tree Fall 56 

13. Site 38BK75 Site Plan 58 

14. Site 38BK75 Test Unit 1 60 

15. Site 38BK75 Profiles Test Units 1-1A 61 

16. Site 388K76 Testing Site Plan 62 

17. Photos of Site 38BK76 and Site 38BK245 Testing 63 

18. Site 38BK76 Test Unit 3 65 

19. Site 38BK75 and Site 38BK76 Relative Amounts of 

Nonlocal and Local Ceramics 70 

20. Chains of Title for Yaughan and Curriboo Plantations 77 

21. 1737 Plat of Yaughan Plantation 78 

22. 1862 Plat of Curriboo Plantation 80 

23. Cordes Geneaology 82 



24. Photos of Field Methods at Sites 38BK76 and 

Site 38BK245 90 

25. Site 38BK76 Collection Strategy 91 

26. Site 38BK75 and Site 38BK76 Site Plan Yaughan 

Plantation 99 

27. Site 38BK245 Site Plan Curriboo Plantation 100 

28. Typical Trench Cross Section 101 

29. Site 38BK75 Site Plan 105 

30. Site 38BK75 Block Excavation 106 

31. Site 38BK75 Block Profiles • Ill 

32. Photos of Site 38BK75, Structures 75Bi and 75B 2» 

Site 38BK76, and Structure 76A 112 

33. Site 38BK75 Feature F5 Distribution of 

All Artifacts 115 

34. Site 38BK75 Feature F5 Distribution of All 

Architecture Group Artifacts 116 

35. Site 38BK75 Feature F5 Hypothetical Wall at 

Structure 75B 2 .117 

36. Site 38BK75 Feature F25 119 

37. Site 38BK75 Feature F2 120 

38. Site 38BK75 Feature F22 121 

39. Site 38BK75 Features F26, F27, and F28 122 

40. Site 38BK75 Feature F29 124 

41. Site 38BK75 Feature F33 126 

42. Site 38BK75 Features F30, F31, and F32 127 

43. Site 38BK76 Plans of Structures 76C, 76D, 76M, 

and 76E and Associated Features 137 

44. Site 38BK76 Feature F8 .142 

45. Site 38BK76 Plans of Structures 76F, 76G, 761, 76A, 

and 76H 144 



46. Site 388K76 Feature F12 146 

47. Site 38BK76 Feature F13 147 

48. Site 38BK76 Structure 76A Block Excavated 148 

49. Site 38BK76 Structure 76A Block Excavated Profiles . . . .150 

50. Site 38BK76 Features F33 and F82 151 

51. Site 38BK76 Plans of Structures 76K and 76L and 

Associated Features 152 

52. Site 38BK76 Plan of Structure 768 154 

53. Site 388K76 Structure 76B^ and 76B2 Block Excavation . . .155 

54. Photos of Structure 76B and Structure 245B 156 

55. Site 388K76 Structure 768 Block Excavated Profiles . . . .157 

56. Site 38BK245 Structures 245B and 245G 169 

57. Site 388K245 Plan of Structure 245H 172 

58. Site 38BK245 Feature F12 174 

59. Site 38BK245 Plan of Structure 245D 175 

60. Site 38BK245 Plan and Profile of Structure 245A 177 

61. Photos of Structure 245A and 245K 178 

62. Site 38BK245 Features F5 and F6 180 

63. Site 38BK245 Feature F63 Plan and Profile 183 

64. Site 38BK245 Structure 245K Kiln Plan View 184 

65. Site 38BK245 Structure 245C Block Excavation 186 

66. Site 38BK245 Structure 245C Block Excavation 187 

67. Photos of Structure 245C and Chimney 188 

68. Site 38BK245 Structure 245 Chimney Foundation 

Fill Profile 189 

69. Site 38BK245 Structure 245C, Cellar Profile 192 

70. House Orientation at Yaughan and Curriboo 

Plantations 198 

71. Building Unit and Structure Areas 204 



72. Slipware Seriation 224 

73. Colono Pots 227 

74. Colono and Catawba Rim Forms 228 

75. Decorated and Smoothed Colono 230 

76. Colono Bases, Appendages, and Objects 231 

77. Catawba and Other Kitchen Group Artifacts 235 

78. Catawba Pitcher 236 

79. Seriation Based on Nonlocal and Local Ceramics 244 

80. Seriation Based on Rim Counts 245 

81. Activity Group Artifacts 257 

82. Clasp Knives 258 

83. Activity Group Artifacts 259 

84. Chest Lock 260 

85. Padlock 261 

86. The Mean Ceramic Dates and Mean Occupation Dates 

for Site 38BK76 and Site 38BK75 281 

87. Number of Transactions by Month 307 



LIST OF TABLES 

1. Population of South Carolina and St. Stephen's 

Parish, 1790-1850 31 

2. Population of South Carolina, Berkeley County and 

St. Stephen's Parish, 1890-1970 32 

3. South Carolina Slave Population, 1703-1724 33 

4. Population Growth, St. Stephen's Parish, 1790-1850 36 

5. Frequency Distribution, Slave Ownership, 

St. Stephen 's Parish 

(a) 1840 38 

(b) 1850 38 

(c) 1860 38 

6. Berkeley County, Cotton and Corn Production 1889-1939 ... 41 

7. Historic Ceramics Recovered from Site 38BK75- 

Site 38BK75-Locus B (Testing) 66 

8. Historic Ceramics Recovered from 

Site 38BK75-Locus A (Testing) 68 

9. Historic Ceramics Recovered from Site 388K76 (Testing) ... 69 

10. Site 38BK75 Structures 107 

11. Site 38BK75 Features 108 

12. Site 38BK75 Artifact Patterns by Structure 

and Feature 109 

13. Site 38BK76 Structures 130 

14. Site 38BK76 Features 132 

15. Site 38BK76 Artifact Patterns by Structure and Feature . . .134 

16. Site 38BK245 Structures 160 

17. Site 38BK245 Features 162 

18. Site 38BK245 Artifact Patterns by Structure 

and Feature 164 

19. Revised Cellar Artifact Pattern Structure 245A 179 

20. Posthole Distance Data, All Trench and 

Posthole Structure 195 



21. Site 388K76 Structure Groups A and B Kitchen 

Artifacts 199 

22. Site 38BK76 Structure Groups A and B Artifact 

Patterns 200 

23. Site 388K76 Structure Groups A and B Kitchen 

Artifacts 201 

24. Site 388K76 Structure Groups and Various Sites 

Ceramics and Glassware 201 

25. Chi-square Values of Ceramics and Glassware 202 

26. Widths, Lengths, and Axes of Building Units 205 

27. Various Building Unit Dimensions Compared 207 

28. Average Width/Length Ratios 209 

29. Total Artifacts by Site and Type 213 

30. Total Artifacts for Selected Structures 215 

31. Nonlocal Ceramics 220 

32. Colono Objects 233 

33. Potential Potting Clays 237 

34. Cdlono/Catawba and Nonlocal Ceramics 241 

35. Nonlocal and Local Ceramic Totals 243 

36. Ceramic Rim Counts 246 

37. Correlation Coefficients of Pairs of Forms 247 

38. Kettle Fragments and Colono Cooking/Storage Forms 249 

39. Bottle Glass Varieties 250 

40. Glass Minimum Vessels 251 

41. Major Glass Forms as a Percentage of Glass Artifacts . . . .251 

42. Major Glass Forms as a Percentage of Total Artifacts . . . .252 

43. Gun Parts and Associated artifacts 253 

44. Modified Pipe Tips at Site 38BK76 254 



45. Tobacco Pipe Stems and Bowls from Excavated 

Contexts Only 255 

46. Measurable Hoes 262 

47. The Frontier Artifact Pattern as Proposed by South 268 

48. Adjusted Fort Prince George Artifact Pattern 

Compared to the Fort Watson Pattern 269 

49. The Artifact Patterns from Spalding's Lower Store 271 

50. Revised Frontier Artifact Pattern Model 271 

51. Artifact Patterns from the Delaware State House 

Excavations 272 

52. Artifact Patterns from the Hepburn-Reonalds House 

and Camden Toft 8 274 

53. The Public Interaction Pattern 275 

54. The Public Interaction Pattern Adjusted to Exclude 

Spalding's Lower Store 276 

55. Revised Carolina Artifact Pattern Observed 

Range and Mean 278 

56. The Revised Carolina Artifact Pattern 279 

57. The Carolina Slave Artifact Pattern 283 

58. Comparison of the Carolina Artifact Pattern and 

the Carolina Slave Artifact Pattern 285 

59. Soil Acidity 287 

60. Seeds by Site 289 

61. Botanical Material by Weight 289 

62. Animal Food Sources 293 

63. Minimum Number of Individuals 294 

64. John Cordes Estate, Account Book 305 

65. Incidence of Bonds among Participants 

in the Exchange 305 

66. Cross Tabulation, Selected Items by Individual /Firm . . . .305 

67. Goods and Services Supplied Henry Ravenel 308 



68. Transmission of Slave Property by Inheritance 310 

69. Bills of Sale and Mortgage 317 

70. Slave Demography in Cordes' Inventories 319 

71. Mean Slave Household Size and Distribution of Slave 

Households by Size 322 

72. Artifact Types Compared 333 



I. INTRODUCTION 

The Cooper River Historic Sites Investigation began with testing and histori- 
cal research on five sites in Spring, 1979. A sixth historic site, 38BK245, 
had already been tested by the staff of Interagency Archeological Services- 
Atlanta and had been included among the mitigation priorities. The testing 
strategy, described in detail in the report which follows, involved surface 
reconnaissance, metal detector survey, and placement of three foot square 
excavation units. Two sites were eliminated during the testing phase due to 
extreme disturbances from agricultural plowing or disturbances related to 
plantation pine plantings (Garrow and Wheaton 1979). A third site was found 
during mitigation to have been noncultural in origin and was abandoned after 
a week of excavation. 

The primary excavation effort was centered on three numbered sites (388K75, 
388K76, and 38BK245) and extended for six months in the field with a six 
person crew (Figure 1). Two of these sites, 38BK75 and 38BK76, turned out to 
be slave occupations associated with Yaughan Plantation, while the third, 
388K245, was a portion of adjacent Curriboo Plantation. 

The initial testing project was conducted under a relatively flexible re- 
search design. The project Request for Proposal included a series of guiding 
questions that were to be used to organize research on the sites. Those ques- 
tions were grouped into three categories that were utilized in the proposal. 
The first category dealt with the ethnic identity and social context of the 
people who produced the archaeological remains under study. The second cate- 
gory of questions then placed the individual sites within their larger re- 
gional context, changing patterns of settlement through time, and economic 
variables. The third category of questions sought to place the sites within 
the larger realm of historical archaeology. 

The testing research design elaborated on the guiding questions in the Re- 
quest for Proposal through suggesting that the major thrust of the testing 
would be to determine if proper conditions existed on the sites to extract 
useable artifact patterns following South (1977a). The artifact pattern ap- 
proach was perceived as the manner 1n which the questions of ethnicity and 
comparability could be addressed. The. primary criteria established in the 
proposal for determining if potential for patterning data existed was the 
presence or absence of architectural remains in association with trash dis- 
posal. It was believed (and correctly as events transpired) that the key to 
constructing meaningful artifact patterns rested with knowing exactly where 
one was excavating within a site and insuring that individual structures and 
their associated domestic trash were included in the excavated areas. The 
priority for the testing phase then became a question of determining if the 
Individual sites contained architectural remains that were sufficiently pre- 
served to justify full scale mitigation. 

The testing phase included historical research that was Intended to answer 
basic questions concerning the sites under investigation. Four goals were 
identified in the proposal. Those goals were: 

1. Development of a concise economic history of the Huguenot settlement in 
the general project area. 



2. Construction of chains-of-title for the parcels that contain the testing 
units. 

3. Compilation of comparative economic data from an as yet undetermined Bri- 
tish Colonial /Anglo-American plantation settlement area. 

4. Identification of diaries and personal papers generated by individuals 1n 
the general project area that would provide insights into day-to-day life 
within local plantation systems. 

Unfortunately, it proved to be impossible to do more than construct the 
chains-of-title during the historical research in the testing phase. It was 
assumed during the preparation of the proposal that the previous research 
(White n.d.) conducted on the historic sites would be useable with little 
more than verification. Attempts to cross-check that earlier research re- 
vealed that through a map error the research had centered on the wrong plan- 
tations. This meant that the limited research time available during the 
testing phase had to be employed in pursuing the item of highest priority, 
the chain-of-title. 

The Request for Proposal emphasized that the major properties under study, 
Sites 38BK75, 38BK76, and 388K88 had been owned and occupied by Huguenots. 
The research design amplified that emphasis, and early in the project it was 
assumed that evidence of ethnicity would relate to Huguenot versus Anglo- 
American occupation. It was not known prior to the testing phase if the 
sites had been occupied by owner, overseer, or slave. Also, the historical 
research prior to testing was not adequate to indicate that the Huguenot 
descendants who owned the sites were sufficiently removed from France in both 
time and space that they had become archaeological! y Indistinguishable from 
their Anglo-American neighbors. That information was to come much later in 
the project, as was the revelation that the sites that ultimately came under 
mitigation were primarily the products of occupations by Afro-American 
slaves. 

The testing strategy used during the fieldwork was successful in delineating 
the mitigation priorities, but was not totally responsive to the types of 
sites present. The strategy involved surface reconnaissance, metal .detector 
survey, and excavation of three foot test units. Architectural remains or 
features were found on two sites, while a third site yielded a mysterious 
feature that later proved to be nothing more than a burned tree trunk. The 
problem encountered with the testing strategy centered around the size of the 
excavation units. Three foot squares were simply too small to reveal the 
type of architectural evidence sought. The metal detector survey did locate 
the structural areas accurately as metallic clusters, but the units were, in 
most cases, too small to locate postmolds or allow interpretation of the hi- 
therto undescribed wall trenches that made up the 1n-the-ground evidence of 
the structures. The test units did allow for an assessment of prior impacts 
and historic soil loss, but only hinted at the amount of architectural re- 
mains present at Sites 38BK75 and 38BK76. 



The report prepared at the close of the testing phase consisted of documented 
recommendations covering the five tested sites as well as Site 38BK245 that 
had been tested by Interagency Archeological Services-Atlanta. Sites 38BK75, 
388K76, and 38BK73 were recommended for mitigation under various levels of 
effort. In addition, an 11-week field effort was recommended for Site 
38BK245. Through subsequent negotiations, the level of effort on all of the 
sites save 388K73 was increased. The historical research was also increased 
in scope and funding. 

A number of problems impeded the field phase. The field crew was moved from 
site to site as construction priorities changed, resulting in time lost due 
to the inefficiency of gear ups and gear downs. A portion of Site 38BK76 
scheduled for block excavation was damaged by a clearing contractor who ig- 
nored the flagging around the site. Also, Site 38BK76 turned out to be much 
larger than anticipated and the actual boundaries of the site were determined 
in the mitigation process. The inadequacies of the testing approach created 
problems during the entire field effort on Sites 388K75, 38BK76, and 38BK245. 
The investigators were quite confident that the first two sites were well 
preserved and contained numerous structures based on the soil profiles and 
metal detector survey. Concrete proof of that feeling was lacking until the 
mitigation effort was underway on each site, and the fieldwork became a con- 
tinuous round of increased levels of effort pieced onto previous work. The 
piecemeal approach used on these sites resulted in a level of inefficiency 
that diminished the outcome of the research. 

Site 388K245 presented special problems. Use of part of the site as a borrow 
pit had apparently destroyed the main house and an unknown portion of the 
slave quarters. The boundary of the Corps of Engineers property ran through 
the site and investigation on non-Corps land was forbidden. Also, the test- 
ing conducted by IAS-Atlanta at the brick kiln had been done the previous 
winter. Between the stripping of the main portion of the site prior to 
IAS-Atlanta involvement and the actions of freezing and thawing, a 
significant portion of the site had been lost before mitigation. Damage 
varied over the site and ranged from an almost total loss of the surface 
brick kiln to moderate to heavy damage on a cellar and a series of 
structures. Site 388K245 was the last site to be completed and the clay 
subsoil was baked to the point that some of the features could not be exca- 
vated utilizing sensitive techniques. Significant data was extracted from 
38BK245 despite these problems, but Information was certainly lost on the 
site from prior impacts. 

The Corps property boundary proved to be a problem on Sites 38BK75 and 38BK76 
as well as on 388K245. An area assumed to contain the main house for the 
slave quarters of 388K75 and 38BK76 was located outside of the Corps' pro- 
perty and thus could not be investigated during the mitigation. An extensive 
surface collection was made of that area by the crew working on their own 
time, so that at least in this instance, some information was recovered. 
There is little doubt that the results achieved concerning Yaughan Plantation 
(Sites 38BK75 and 38BK76) would have been greatly enhanced if more comparable 
data could have been extracted from the main house area. 



The historical research continued during the mitigation fieldwork and gradu- 
ally pieced together a picture of the historical occupations at Yaughan and 
Curriboo Plantations. The last of the archival data collection and analysis 
was not completed until well into the archaeological analysis phase. 

It was recognized fairly early in the testing phase that Sites 38BK75, 
388K76, and 38BK245 were primarily comprised of Afro-American slave domestic 
occupations. A level of confirmation of that impression was given by the 
historical research, and by the close of the fieldwork, there was no longer 
any doubt concerning the nature of the sites. The Cooper River Historic 
Sites Investigation was undertaken through a relatively new contracting pro- 
cedure whereby the analysis phase of the project was negotiated as a separate 
contract at the end of fieldwork. This presented the opportunity to refine 
the project research design and lend greater specificity to the guiding ques- 
tions to be asked during the analysis phase. 

Five hypotheses were formulated at the conclusion of the fieldwork to insure 

orderly progress during the analysis. Each hypothesis was adequately tested 

during the analysis phase, and each was explored in terms of both archaeologi- 
cal and historical implications. 

The first hypothesis dealt with the basic nature of colonoware (Ferguson 
1977) ceramics. 

Hypothesis 1 . The Colono ceramics recovered from Sites 
3bBK/b, 3tfBK/6, and 388K245 represent ceramics that were made 
by slaves who occupied the plantations, and that the slaves 
produced those wares for their own use. It is further hypoth- 
esized that the Yaughan (and) Curriboo Plantation samples are 
representative of the colonowares that were being made and 
used by African slaves In coastal South Carolina during that 
period. 

Three potential variables were considered while testing Hypothesis 1. The 
first variable was that the colonowares extracted from the sites may not be 
truly representative of ceramics being produced and used by African slaves 1n 
other areas. The first variable was controlled through limited comparison of 
the ceramic samples from Yaughan and Curriboo with collections from other 
sites. The second variable considered under Hypothesis 1 was that at least 
some of the slaves on the plantations were Indians and that the ceramics pres- 
ent were produced by Indians Instead of Blacks. Careful historical research 
was employed to control that variable. The third variable considered was 
that the Afro-American slaves were the products of a number of different cul- 
tural groups, and that those diverse backgrounds would be reflected in a high- 
ly heterogeneous ceramic assemblage. That variable proved to be impossible 
to control with available data, and although there appears to be a degree of 
homogeneity in the ceramic sample, 1t still is not possible to attribute the 
ceramic styles to specific African groups. 



The second hypothesis was largely an outgrowth of research by Lees and Kimery- 
Lees (1978) at Limerick Plantation in coastal South Carolina. Those investi- 
gators observed that there was a relative decline in the use of colonoware 
after 1775. Yaughan and Curriboo Plantations offered a good opportunity to 
study the nature of the decline of colonoware, and Hypothesis 2 was formu- 
lated to facilitate that effort. 

Hypothesis 2 : Colonoware declined in importance at the 
plantations as time passed. Conversely, there was a trend 
toward greater dependence on non-1 ocally produced ceramics 
from the eighteenth to nineteenth centuries. 

The single variable identified that could affect Hypothesis 2 dealt with the 
nature of use of colonoware ceramics. That is, it was not known at that 
point in the analysis if the relative frequencies of colonoware through time 
was a function of chronology, status, acculturation, or a combination. That 
variable was controlled satisfactorily during the analysis. 

The third hypothesis was the primary organizing question used throughout the 
project. That hypothesis dealt with the concept of artifact patterning as 
advanced by South (1977a) and was viewed as the major vehicle for cross-site 
comparisons. 

Hypothesis 3 : Patterns of artifact disposal on archaeological 
sites are culturally determined and can be discerned through 
careful analysis. Since African slaves were the products of 
radically different backgrounds than Anglo-Americans, the 
pattern of artifact disposal on African slave domestic sites 
discernible from that present on Anglo-American 
difference will be expressed in varying frequen- 
artlfact categories that the sites share in com- 
as the absence of certain categories. 

Two potential variables were anticipated in the case of Hypothesis 3. First, 
it was anticipated that the concept of patterning as put forth by South 
(1977a) and Lewis (1973) may not reflect more than a few site specific cases. 
The consistent nature of the artifact patterns achieved on the Cooper River 
Historic sites seem to indicate that that variable was not at work in the 
case of this investigation. The second potential variable noted was that the 
artifact patterns could be influenced by the degree of acculturation present. 
That is, it was considered possible that Afro-Americans could have been so 
thoroughly acculturated at the time of occupation of the sites that no differ- 
ences in artifact pattern would exist between them and Anglo-Americans. 
Significantly, the artifact pattern results do seem to indicate that accultur- 
ation was progressing through time on the sites and that the beginnings of a 
transition to artifact patterns more nearly like those on British-American 
sites was observable in the later occupations. 

The final two hypotheses offered explanations for the architectural shifts or 
differences noted on the sites during the excavation phase. 



should 


be 


sites. 


The 


cies of 


the 


mon, as 


well 



Hypothesis 4 : The earliest construction technique in use 
within the Vaughan Plantation slave quarters involved wall 
trench construction coupled with the use of a few individual 
postholes. As time passed, that mode was superceded by the 
use of entirely individually set posts. This construction 
sequence represents a transition from construction techniques 
used in the Carribbean and/or Africa and reflects the greater 
acculturation of slaves into the British-American sub-culture 
as time passed. 

Hypothesis 5 : The earliest construction technique within 
Curriboo Plantation involved wall trench construction coupled 
with the use of a few individual postholes. As time passed, 
that mode was superceded by the use of brick pier construction 
for major plantation outbuildings. This construction sequence 
represents a transition from a frontier pattern to a more per- 
manent settlement pattern and a successful adaptation to pre- 
vailing economic trends. 

Hypothesis 4 dealt exclusively with domestic architecture, while Hypothesis 5 
addressed the non-domestic architecture encountered at Curriboo Plantation. 
The major variable Identified In the research design that could affect those 
two hypotheses dealt with the nature of French architecture at the time the 
sites were occupied. It was recognized that wall trench construction was 
known on some French occupied sites, and 1t was thought possible that the 
wall trenches encountered at Yaughan and Curriboo could relate to that tradi- 
tion. That variable is discussed In the architecture section of the 
following report. 

The historical research proposed for the analysis and reporting phase was 
viewed as an adjunct to pursuing the stated archaeological goals. A series 
of questions concerning the chronology, social content, and economic system 
of the plantations were posed. The historical research phase was unusually 
productive in this regard. Despite the fact that personnel changes occurred 
during the research and synthesis stage and the funding for the historical 
research was not continuous, an excellent data yield was achieved. Perhaps 
the key Information developed during the historical research was the fact 
that demographic continuity existed at both plantations within the slave 
populations during the course of the occupations. This means that the 
archaeological data retrieved was produced by relatively stable communities 
and that changes reflected 1n the archaeological record were indicative of 
cultural /social versus population changes. This validates the concept that 
the archaeological research did measure acculturation as reflected by mate- 
rial culture for the period from approximately 1740 to 1826. 

The sub-specialty of Afro-American archaeology has an unfortunately brief his- 
tory. While the first substantive research conducted on a site occupied by 
an Afro-American was published 1n 1945 (Bullen and Bullen 1945), it was not 
until the late 1960s and early 1970s that a systematic attempt was made to 
develop Afro-American archaeological research (Ascher and Fairbanks 1971; 
Fairbanks 1972; and Schuyler 1974). The most extensive research conducted at 



that time centered on the Florida and Georgia coastal areas and dealt 
exclusively with nineteenth century sites. While the coastal investigations 
conducted by the University of Florida appear to have begun as an attempt to 
identify surviving African traits in the archaeological record (Fairbanks 
1972:90), those investigations have since been broadened to incorporate a 
number of other anthropologically sound research questions. Perhaps the 
major contribution of the Florida studies to date has been through the work 
of John Otto (1975, 1977, and 1979). Otto's dissertation (1975) remains the 
most exhaustive work on a single plantation site that has been reported under 
the rubric of AfroAmerican archaeology. The Florida program in Afro-American 
archaeology is still underway (Mull ins-Moore 1979 and 1980; Singleton 1979; 
and Hamilton and Marrinan 1979) and promises to make additional substantive 
contributions to our knowledge of nineteenth century Afro-American life. 

Research on Afro-American sites was not restricted to the Florida program dur- 
ing the 1970s. Perhaps one of the most influential projects to the develop- 
ment of Afro-American archaeology was the work undertaken by James Deetz 
(1977:138-154) in 1975 at the Parting Ways Site near Plymouth, Massachusetts. 
Deetz 's primary emphasis paralleled that of Fairbanks (1972) in that he was 
interested in survivals of African traits on a late eighteenth to mid-nine- 
teenth century site, but he also focused attention on the whole question of 
Afro-American archaeology through his eloquent popularized account. 

The New York City area witnessed some of the earliest sustained archaeologi- 
cal work on Afro-American sites. Robert Schuyler (1974) conducted historical 
and archaeological work at Sandy Ground on Staten Island as early as 1971, 
but, to date, a comprehensive report on that project has not been published. 
The available published sources on that work indicate that once again the 
sites investigated were relatively late, with the Afro-American settlement at 
Sandy Ground postdating 1820 (Schuyler 1974:18). 

The Weeksville Project, located in the Bedford-Stuyvesant section of Brook- 
lyn, was undertaken prior to Schuyler's research, with excavations dating as 
early as 1968 (Sal wen and Bridges 1974; Bridges and Sal wen 1980). Weeksville 
was a black community established around 1827 (Bridges and Sal wen 1980:39) 
and only limited results were gained from the investigation. 

During the late 1970s, interest in Afro-American archaeology spread rather 
rapidly with work conducted in the Southeast (Carillo 1980; Drucker 1979; 
Drucker and Anthony 1979; Lees 1979; Lees and Kimery Lees 1978; Lewis and 
Haskell 1980; and Smith 1976), the Mid-Atlantic (Geismar 1980, Kelso 1976; 
and Outlaw, Bogley, and Outlaw 1977), and the Northeast (Bower and Rushing 
1980). Perhaps one of the major breakthroughs of the late 1970s has been the 
positive identification of ceramic types made by Afro-American slaves. That 
work has been pioneered by Leland Ferguson (1977 and 1980) who has published 
an excellent assessment of "Colono" ceramics. Ferguson's work has keyed 
primarily on coastal South Carolina, North Carolina, and Virginia. The full 
geographical range of "Colono" ceramics 1s currently being explored, but 
available information does Indicate that "Colono" ceramics, presumably of 
African slave manufacture, do occur in the Caribbean (Gartley 1979 and 
Vescelius 1977). 



Although the majority of the work conducted on Afro-American sites in recent 
years has been done following other problem orientation approaches, several 
persistent shortcomings have hampered the development of this sub-specialty. 
Most of the investigated sites have dated from the nineteenth century and 
have dealt with populations who apparently had become rather thoroughly ac- 
culturated within the larger Anglo-American culture. This has made it dif- 
ficult to assess the nature of the ethnic differences that may be apparent in 
the archaeological record. A second difficulty has been the limited scope 
under which most of these projects have been conducted. Many of these 
projects have been little more than extended testing investigations and have 
provided little more than tantalizing glimpses into the nature of Afro-Ameri- 
can sites. A third problem relates to the newness of the sub-specialty. 
That is, much of the research undertaken to date is available in little more 
than summary form as a good bit of the more germane research is still under- 
way or just recently completed. An additional problem, one that has hampered 
many types of historical archaeological research, is the lack of artifact 
quantification in most of the existing studies. Exceptions do exist (Drucker 
and Anthony 1979; Lewis and Haskell 1980), but for the most part reports on 
Afro-American excavations have not Included full artifact lists or descrip- 
tions. 

The investigation of Yaughan (38BK75 and 388K76) and Curriboo Plantations 
(388K245) should fill some major gaps in Afro-American archaeology. Both 
sites were apparently occupied by 1740, and the latest portion of Yaughan was 
abandoned by approximately 1826. Large scale excavations were undertaken on 
both sites, with numerous structures fully Investigated. Perhaps the most 
important revelation achieved from the fieldwork and analysis was the nature 
and extent of acculturation evident from the earlier to later portions of the 
sites. This is hardly surprising given the early dates of the occupations, 
but 1t is heartening to note that fairly subtle culture change can indeed be 
detected through archaeological Investigation, especially when coupled with 
sound historical research. 



10 



II. ENVIRONMENTAL SETTING 

Introduction 

The Cooper River Rediversion Canal Project is located in the northern third 
of Berkeley County, South Carolina, between Lake Moultrie on the west and the 
Santee River on the east (Figure 2). Berkeley County is situated in the 
Atlantic Coastal Plain Province (Brockington 1980:5), sometimes termed the 
Atlantic Coast Flatwoods area (Long 1980:47). This physiographic province is 
characterized by a low flat topography, cultivated fields, and pine and 
oak-hickory woodlands. In general, it tends to be swampy and poorly drained. 
The total environment affected not only prehistoric settlement, but also 
historic settlement in the area, much more than is the case today. For this 
reason, a general physical description of Berkeley County and the sites 
discussed in this report is necessary. 

Topography 

The geologic formations underlying Berkeley County and which have most great- 
ly affected the soils are unmetamorphosed sedimentary sandstones, limestones, 
and shales. The uppermost formations, having the most direct impact on soil 
building processes, are Black Mingo sandstones and limestones (Brockington 
1980:5). The large number of "orthoquartzite" prehistoric artifacts attests 
to the utility of this sandstone formation for tool making by prehistoric 
peoples, although the quality of the material does not approach that of 
coastal plains chert. 

Overlying the sandstone and limestone formations are marine and fluvial soils 
deposited during the Pleistocene. These soils were deposited in four ter- 
races: the Wicomico, 70-100 feet A.S.L.; the Penholoway, 42-70 feet A.S.L.; 
the Talbot, 25-42 feet A.S.L.; and the Pamlico, 0-25 feet A.S.L. (Long 1980: 
43). Based on associated soil types and elevation, Sites 38BK73, 38BK75, 
38BK88, and 38BK91 were located on the Wicomico terrace; 388K245 was on the 
Penholoway terrace; and 38BK76 was on remnants of the Talbot terrace. 

The Santee River forms the northern and northeastern edge of Berkeley County, 
and during the colonial period most settlement was restricted to the south 
bank. The river was navigable by small boats upstream from the Atlantic, at 
least as far as St. Stephen, and provided a means of transport for goods. 
Historical documents tend to show, however, that most transport was over 
roads to Charleston during the eighteenth and early nineteenth centuries 
(Chapter III and Orvin 1973:69). The south bank of the Santee consists pri- 
marily of a wide swamp in the St. Stephen's area. This swamp was utilized by 
some planters for rice cultivation and indications of old rice fields are evi- 
dent on aerial photographs of the swamp near Curriboo plantation (Figure 1). 
No archaeological fieldwork was conducted in this area, however. 

Moving inland from the swamp there is a steep bluff abruptly rising 10 to 20 
feet. It was along this bluff that three of the sites (38BK75, 38BK76, and 
38BK245) were located. According to the Mouzon 1771 Map, most plantations of 
the colonial period were, in fact, located along this bluff. For example, at 
Yaughan and one of the Porcher plantations further upstream, the highest 



12 



knoll or rise near the bluff was apparently occupied by the main house, and 
the outbuildings and slave quarters are on lower areas and often closer to 
the river. 

Inland areas tend to be nearly level with low rises. Inland low spots fre- 
quently contain standing water much of the year, and the larger of these 
areas are usually termed bays, a synonym for swamp, and are included on maps 
of the area as important topographic features. Sites 38BK73, 38BK88, and 
388K91 were located in this zone. 

The project area can, therefore, be divided into three zones: the Santee 
River floodplain, the bluff, and the inland zone. Only the last two contain 
habitation sites with structural remains in the project area. As noted be- 
low, soils, fauna and flora tend to follow these zones, with some overlapping 
in specifics. 

Soils 

The following information on Berkeley County soils is taken from Long (1980), 
the Soil Conservation Service's county soil survey. For a more in-depth dis- 
cussion of general soil types in other parts of the county, this work should 
be consulted. 

Site 388K73, a field scatter of artifacts and tree fall, was located on 
Bonneau loamy sand with to 2 percent slopes. The site itself was in a 
flat, cultivated field. Bonneau loamy sand is deep and moderately well 
drained. It makes up 2.4 percent of the soils in Berkeley County. It also 
has moderate limitations on agriculture due to leaching of nutrients, and 
fertilizing is therefore needed frequently in small amounts. The main crops 
are cotton, corn, and soybeans. Open and woodland wildlife potential is 
good, but wetland potential is poor. Use as construction material is 
restricted to road fill. Acidity generally ranges from pH 4.5 to 5.5. 

Sites 388K75 and 388K88, a slave quarter and an isolated farmstead, were lo- 
cated on Norfolk loamy sands in level or nearly level cultivated fields. 
This soil type makes up 1.9 percent of Berkeley County and is principally 
under cultivation. It is better drained than Lynchburg (38BK91) and Bonneau 
(38BK73) soils. The main crops cultivated on this soil- are tobacco, cotton, 
corn, and soybeans. The only real problem for agriculture is erosion, which 
can easily be controlled by windbreaks, crop rotation, and contour plowing on 
more sloping land. It has good potential for open and woodland wildlife, but 
very poor potential for wetland species. It 1s reasonably good for road 
fill, but unsuited for construction sand. Acidity is not as great a problem 
with Norfolk soils as with the other soils discussed here, having a pH rang- 
ing from 4.5 to 6.0. It might be noted that none of the soils approach 
neutral pH any closer than the 6.0 recorded here. 

Site 388K76, a slave quarter, had Pantego fine sandy loam and was located in 
a mature oak-h1ckory forest. Pantego soils are nearly flat, deep, and ^ery 
poorly drained. These soils make up 3.6 percent of county soils. Most Pante- 
go soils are 1n woodland or are occasionally used for pasture. Poor drain- 
age, combined with low organics, present moderately severe hazards to agri- 
culture. Wildlife potential is fair for openland species and good for 



13 



woodland and wetland species, the only soil discussed here which is compati- 
ble with wetland species. Pantego loam is unsuitable as a construction ma- 
terial and is relatively acid, pH 3.6 to 5.5. 

Site 38BK91, the scattered remains of a brick clamp, had Lynchburg loamy 
sands. This soil type occurs on 0-6% slopes, and the site itself is on 
generally flat low ground. Lynchburg loamy sand makes up 3.6% of Berkeley 
County and is principally in forest. Lynchburg sands have moderate limita- 
tions on agriculture due to wetness which hinders plant growth. Drainage and 
frequent plowing are required to dry out the soil and maintain tilth. It has 
good potential for open and woodland wildlife, but only fair for wetland 
species. It is not of particular use for construction, such as road building 
or sand 1n concrete. It also tends to be quite acid, pH 3.6 to 5.5, which 
partially explains its wide use for pine plantation. 

Site 388K245, a portion of Curriboo Plantation, was situated on Wahee loams, 
which are nearly level, deep, and poorly drained. These soils represent 6.2 
percent of soils in the county and are mainly in pine forest. Poor drainage 
greatly hinders agriculture. With drainage these soils are used mainly for 
com, soybeans, and cabbage. Wildlife potential is good for open and wood- 
land species and only fair for wetland species. Construction use is very 
limited and acidity 1s between pH 4.5 and 5.5. 

Climate 

The climate of Berkeley County can be characterized as subtropical. Summers 
are long, hot, and humid. The warmest month is July with an average tempera- 
ture of 80.1°F., and the coldest month is January with 46.8°F. Two years in 
ten (as occurred during the fieldwork), July will have a temperature higher 
than 99.1°F. July is also the wettest month, averaging 7.1 inches of rain- 
fall (Long 1980:93). Rain at this time of the year comes in afternoon 
thundershowers. During the summer of 1979, these thunderstorms were very 
regular, arriving out of the southeast between 3 and 4 o'clock for days on 
end. The driest month is November with 2.0 inches of rain. There is an 
average of 260 frost free days a year in Berkeley County, beginning in early 
March and continuing until early November. Sixty-six percent of the yearly 
rainfall comes between April and September, so that this overlap results in 
an excellent climate for single or successive planting of crops. 

Habitats 

The native fauna of Berkeley County can be divided into three groups: open- 
land, woodland, and wetland habitats. These three groups have already been 
mentioned in the soil descriptions above. Openland fauna include rabbit, 
fox, and assorted gamebirds, including quail and dove. The woodland habitat 
attracts turkey, opossum, fox, raccoon, deer, and bobcat. Wetland species 
represented are duck, goose, heron, muskrat, mink, and occasionally alligator 
(Long 1980: 36). During the course of fieldwork, quail, dove, wild turkey, 
deer, and heron were noted along with water moccasins, rattlesnakes, and 
other non- poisonous snakes. Wild fauna seemed abundant, especially in and 
near wooded areas. Archaeologlcally, white- tailed deer, goose, opossum and 
snakes were found at Yaughan and Curriboo in trash features. 



14 



These animals feed on and are provided with cover by a wide variety of wild 
and domestic plants. In open areas, wild plants consisted of Indian grass, 
goldenrod, and pokeweed as well as other weeds and grasses. Wooded areas 
were generally mature oak-hickory or pine forests with minor amounts of sweet- 
gum, dogwood, briars, shrubs, and vines. The only site approaching a native 
habitat before testing was 38BK76, which had been covered by an oak-hickory 
forest with occasional pine and dogwood. Sites 38BK73, 38BK75, 388K88, and 
388K245 were cultivated fields before testing and 38BK91 was in pine planta- 
tion. 

Natural Resources 



From the archaeological and historical evidence the natural resources used by 
the historic inhabitants for commercial purposes were relatively restricted. 
There are no precious metals or commercially exploitable minerals, except 
lime, available to the occupants in Berkeley County. On the other hand, there 
were a few other natural resources which provided supplementary sources of 
income to the plantation owners and to their slaves. 

As land was cleared and before the first commercially successful crops could 
be harvested, the forests provided a major source of outside income for many 
planters. At Site 388K245, a warehouse or processing shed cellar had a thick 
layer of pine tar which indicated storage or production of naval stores. 
This will be discussed in later sections of this report. No indications were 
found archaeological ly of lumber production at any of the sites, but histori- 
cal documents indicate that pine pitch was being traded between planters for 
farm use (Chapter III), and general historical accounts point out that the 
region was known for production of naval stores. Wood products in the form 
of lumber and barrel staves were also sold to factors in Charleston for even- 
tual shipment and sale to Britain during the eighteenth century (Orvin 
1973:58, Sirmans 1966:226). 

Another major resource was the soil Itself. This is evident from the archaeo- 
logical record in two forms. Sites 38BK91 and 38BK245 had remains of brick 
clamps. Without trace element analysis, it is impossible to be completely 
certain that the source for brick clay was the St. Stephen area. However, ex- 
amination of soil samples and historical documents strongly supports the con- 
tention that brick making was a local industry using local materials (Dubose 
n.d.). The red clay subsoils, especially near the brick clamp at 388K245, 
fire to the same consistency, color range, and hardness as bricks from the 
clamps and from the other sites excavated. Evidence for locally made brick 
is rife in the literature (Noel Hume 1978). Its presence in St. Stephen 
would not be surprising. The economic importance of brick making for the 
sites discussed here is not clear, but indications are that the use of brick 
structures and, therefore, the demand for brick, were not great. 

Of more economic and soda! importance was the use of local clays for the pro- 
duction of unglazed ceramics. No mention 1s made of slave produced ceramics 
1n day book accounts accessed here, but the vast quantities of colonoware re- 
trieved from nearly all of the sites points to its importance in the daily 
life of the slave population. At Site 388K76, over 88 percent of the ceram- 
ics and over 67 percent of the total artifact assemblage were colonoware. 
This site is the most extreme example, but significant amounts of colonoware 



15 



were also recovered at 38BK75 and 38BK245. A much lower percentage of colono- 
ware at later and non-slave sites indicates that its primary importance was 
within the material culture of the slave population. As a result of this 
abundance of colonoware, the savings felt by the slave owners, who did not 
.i».ed to supply slaves with more costly non-local ceramics, must have been 
significant; whether or not they were actually aware of it. It is possible 
that some slave owners would not have made up for a lack of colonoware by 
some other means (for example, tin plates, non-local ceramics, wooden 
trenchers), but on the whole, directly or indirectly, slave owners profited 
by colonoware production. 

Other natural resources provided by the environment include edible plants and 
animals. Many of these have already been mentioned. A complete discussion 
of subsistence is presented in a separate section. It is enough to say here 
that archaeological evidence for use of native resources for food is poor. 

An historical account of agriculture is presented elsewhere. Of interest to 
this discussion are requirements of indigo, rice, and other staple crops with 
respect to the available natural resources. The indigo plant requires a light 
sandy soil, complementing rice, which requires a wet swampy environment (C.W. 
1755:202, de Beauvais 1769). Both types of environment are present in the 
study area. R1ce and indigo also complement each other with respect to 
seasonality (Sirmans 1966:269), thus providing continuous, nearly year round 
employment for the work force, without the need of resorting to forest pro- 
ducts or brick making for outside income. 

From deeds, plat maps, and modern soil maps, it appears that both Yaughan and 
Curriboo plantations consisted of approximately 50 percent swampland, of po- 
tential use for rice cultivation, and 50 percent upland of use to habitation 
and upland crops, including indigo. This would mean that approximately 500 
acres at Yaughan and 600 acres at Curriboo were available for rice, and the 
same amounts for indigo and other crops. According to historic sources, one 
to four acres of Indigo was the average cultivated per field hand (C.W. 
1755). Assuming that an average of 30 field hands (out of a total 40 hands 
at Yaughan) were available for fieldwork, a maximum of 120 acres could have 
been planted in indigo at anyone time, leaving the remainder for other uses. 

Unfortunately, the available records cannot tell us how close the owners came 
to this hypothetical goal for indigo production. This is so because the 
available records are incomplete, entries concerning production may not repre- 
sent yearly totals, entries may show only John Cordes' portion of respective 
harvests, and they may represent only that portion of produce sold on the 
open market for cash. With more complete and reliable data, it might be pos- 
sible to calculate not only the economic importance of indigo to the planta- 
tion, but also the numbers of slaves and acres of land devoted to indigo, 
rice, and other crops. This kind of data would allow detailed study of land 
use patterns, soil management skills, and labor productivity. 



16 



One conclusion that may be drawn from this data is that not all of the land 
available for indigo and rice cultivation could possibly have been cultivated 
at any one time with the labor available. Whether this was a decision freely 
made by the landowners, or whether economic conditions forced the decision up- 
on them is unknown. But Yaughan with 500 acres and Curriboo with 600 acres 
of uplands, could have employed up to 125 to 150 field hands for indigo culti- 
vation respectively, and all indications are that these totals were neyer 
reached. 

Other crops were also grown on the plantation for home consumption as well as 
for sale or trade in the St. Stephen area. These included grains and vegeta- 
bles which competed with indigo for upland soils. No other crops are report- 
ed to have competed with rice for swampland. As noted above, the soils 
available for cultivation were generally good for corn, tobacco, and cotton. 
Today, soybeans have become one of the major cash crops along with the three 
just listed. 

Intense cultivation of the Santee River floodplain and inland areas since the 
early eighteenth century has had an effect on the topography. This is seen 
mainly by erosion from upland areas and silting of the Santee River flood- 
plain. Silting and flooding are two of the reasons why rice production 
became uneconomical above the tidewater zone at the beginning of the nine- 
teenth century. Today, the floodplain has been left almost entirely to 
natural vegetation and wildlife, with minor attempts at growing upland crops 
in heavily silted portions. Until the past 20 or 30 years, cattle and hogs 
were allowed free rein in the floodplain, according to local informants. 
Today, however, even this use of the floodplain has been curtailed. 



17 



III. HISTORIC OVERVIEW 

Introduction 

The project focused on three historic sites at two plantations. Only a 
minimal amount of prehistoric material was recovered from the sites. For 
these reasons, an overview of the prehistoric period is included as Appendix 
A of this report. The following chapter presents an overview of the develop- 
ment of South Carolina and St. Stephen's Parish with special emphasis on 
economic development and slavery in order to establish a framework for the 
discussions which follow. 

Protohistoric 

The Protohistoric period begins with initial European contact prior to the 
achievement by Europeans of hegemony in this area. It opens in the mid- 
sixteenth century with the Spanish occupation of North America and ends in 
the latter part of the second decade of the eighteenth century with the des- 
truction of the Yemassee. It focuses on the impact these contacts had on the 
indigenous peoples. This section relies heavily on the detailed research 
into Indians of the South Carolina low country completed by Gene Waddell , 
Director of the South Carolina Historical Society in Charleston. 

Waddell (1980) identified 19 tribes Indigenous to the coast between the 
Santee and Savannah Rivers; these included the Escamacu, the Hoya, Edisto, 
Touppa, Mayon, Stalame, Kussah, Kussoe, Wimbee, Combahee, Ashepoo, Wando, 
Sampa, Seewee, Kiawah, Stono, Witcheaugh, Bochiket, and Etiwan. From 1562 to 
1576, the earliest period of European contact in the vicinity of Port Royal, 
no tribes were known to live in the region between Port Royal and Charleston 
Harbor. In the late sixteenth century, however, the coastal tribes, begin- 
ning with the Edisto, who moved from the Broad River- north to Edisto Island, 
began to migrate northward in order to avoid contact with the Spanish (Wad- 
dell 1980: 1-5). After 1580, the distribution of the coastal tribes evi- 
dently remained fairly stable, until the English occupied the area around the 
confluence of the Ashley and Cooper Rivers 1n 1670. Resources were evenly 
distributed, and one section offered little that another did not match. 
Fairly equal size and resources seem to have created a "small-scale balance 
of power" among the Coastal Tribes (Waddell 1980:19). Within their terri- 
tories, the tribes were evidently autonomous and collectively, they were 
disunited (Waddell 1980:16). They seem to have formed no alliances with 
tribes beyond the coastal area, although they appear to have been familiar 
with and known to groups as far south as St. Augustine, as far north as Cape 
Fear, and nearly 500 miles inland (Waddell 1980:22). 

English occupation after 1670, as well as continued Spanish occupation fur- 
ther south, occasioned dislocations of the coastal tribes, which were com- 
pounded by migrations of other tribes into the area. Between 1675 and 1685, 
tribes that had occupied lands .in the vicinity of Charles Town Harbor left, 
or "were removed" from this area; four of these tribes: the Eitwan, Wando, 
Sampa, and Seewee had taken up lands along the Wando River, three miles from 
its mouth. The Seewee continued to move northward, so that by the turn of 
this century, the tribe occupied territory along the coast below the Santee 
River (Waddell 1980:4-5). 



18 



The state of the coastal tribes became more confused as a result of the migra- 
tion of tribes into the area just west of their territories beginning with 
the Westo in the 1660s. English colonists, on their arrival, found the coast- 
al tribes intimidated by the Westo, who were also said to be cannibals. In 
1674, Dr. Henry Woodward visited one of the Westo Towns, which were located 
inland along the Savannah River. Research into the Westo indicates that they 
were a confederation of several peoples, two of which, the Oustacks and Ricko- 
hockans, have been identified in historical sources. The Oustacks were a 
Timucuan people, originally from Guale, who went north in the late sixteenth 
or early seventeenth centuries to evade the Spanish and settle in Virginia 
among the Powhatan confederation. Here, it has been argued, they joined with 
other groups, including the Pickohockans, an Iroquoian people, and the confed- 
eration subsequently drifted south into the Carolinas (Juricek 1964:139, 153, 
156, 160-61). The Westo War of 1680 destroyed the control that the Westo had 
established over the coastal tribes, although the group was still extant and 
believed dangerous when Thomas Newe visited one of their villages in 1682 
(Juricek 1964:160). 

In the 1680s, the Savannah, who appear to have come from the vicinity of Au- 
gusta, replaced the Westo as a buffer between the English along the coast and 
the tribes further inland. Notorious slave hunters who preyed on the weaker, 
coastal peoples, they drifted northward to settle among the Conestoga and 
Delaware in the vicinity of the Susquehanna River (Milling 1969:86-90). The 
Yemassee began to move into the buffer zone in the 1690s and remained signifi- 
cant trading partners until they were destroyed in the Yemassee War of 1715- 
1716. 

In the late seventeenth century, the Cherokee appear to have consolidated 
their control of the Piedmont, forcing the migration of some groups toward 
the coast. Thus, when John Lawson traveled up the Santee River from the 
coast, he went first through the settlement of the Seewee, who had settled 
there from the south and then through the settlement of the Santee, who may 
have originated in areas much further Inland (Waddell 1980:5). Lawson, how- 
ever, traveled through a long stretch of uninhabited land before he arrived 
at the Santee village. The villages of the Seewee are not known to have ex- 
tended further inland than the site of Jamestown, some 25 miles up the Santee 
from the coast, and the first villages of the Santee appear to have been in 
the vicinity of the present Lake Marion (Waddell 1981, personal 
communication) . 

Linguistic analysis indicates that there were at least two groups present 
among the coastal tribes: one Slouan and another, hitherto believed Musko- 
gean and now believed to be an as yet unidentified linguistic group. One of 
the words used to detect the presence of the unidentified language is "Corre- 
boo", the name of one of the plantations under investigation. Both the San- 
tee and Seewee were Siouan, and it 1s possible that members of a tribe asso- 
ciated with the unidentified linguistic group, which concentrated south of 
the Ashley River, may have come up the Cooper during their winter migrations 
to the vacant lands between the Seewee and Santee, which is in the vicinity 
of the present project site (Waddell 1980:33; and personal communication). 



19 



The locations given above are usually the known sites of the summer villages. 
In the fall, the tribes usually divided into smaller groups and moved up- 
stream (Waddell 1980:1). The majority of the tribes depended on agriculture 
for at least half of their subsistence. Hunting, gathering, and fishing 
contributed the remainder, although tribes on the lower coast may have 
depended less upon agriculture than those of the upper coast. European 
encroachment, wars, disease, the Indian slave trade, and developing 
commercial ties that undermined the Indians' self-sufficiency disrupted the 
political, social, and economic traditions of the Mississippian Period. 

Early European Settlement, 1521-1663 

The earliest known contact with the area now called South Carolina was made 
by the Spanish in 1521. On an expedition to capture Indian slaves, Quexos, 
an emissary of All yon, a higher judge from Hispaniola, landed at the mouth of 
Winyah Bay. In 1526, Allyon himself arrived with some 500 people to estab- 
lish a settlement. Early Spanish settlements proved unsuccessful, but their 
attempts were followed by an effort by Jean Ribaut in 1562. Ribaut landed at 
Port Royal Sound with 150 Huguenots and attempted to establish a colony 
there. This enterprise was also short-Hved (Petty 1943:17-18). 

Spain and France both claimed the territory from Cape Fear south to Florida. 
Although rivalry between the two imperial powers was fierce, the Spanish do- 
minated in this area until their Influence in Europe waned in the early sev- 
enteenth century. In 1566, they built a fort on Parris Island, which they 
abandoned ten years later. The Spanish built a second fort on the same site 
in 1577, which they also left ten years afterwards (Petty 1943:18). 

England's claim to this region was based on the voyages of John Cabot. Sir 
Robert Heath's attempt to create a refuge for Huguenots in Carolina under the 
aegis of the English king 1n 1629 came to nothing. The group landed in Vir- 
ginia in 1630 (Petty 1943:18). Domestic troubles and the English Civil Wars 
effectively ended further experiments 1n colonization until Charles II re- 
gained the throne 1n 1660. Compelled to recognize men who had served him 
loyally during his exile, Charles II made generous grants in the New World to 
those courtiers who had participated in his restoration to power. Among 
these were eight men who obtained the Carolina proprietorship in 1663 
(Sirmans 1966:5). 

Population of Colonial South Carolina 

After preliminary voyages and one unsuccessful effort in 1667, permanent set- 
tlement was established in 1670 on the west bank of the Ashley River, across 
from the present site of Charleston. In 1680, the town was moved across the 
river to its modern location on the peninsula between the Ashley and Cooper 
Rivers. Formally named Charles Town, it became the focal point of the grow- 
ing colony (Petty 1943:19). 

Migration to the new colony 1n the seventeenth century derived from four 
sources: the English West Indies, especially Jamaica and Barbados; other 
mainland colonies; the British Isles; and the European Continent, particular- 
ly France. M. Eugene Sirmans ( 1966:4) argued that the migration from 
Barbados and Jamaica was a critical factor in the early development of South 



20 



Carolina. These men brought with them expertise in staple-crop, plantation 
agriculture based on black and Indian slave labor. The earliest commercial 
links were with the West Indian ports (Clowse 1971:163-165), and within a 
decade, the Barbadians, who clustered on Goose Creek, a tributary of the 
Cooper River, organized themselves into a political faction, whose principal 
objective was cornering the Indian trade in deerskins and slaves (Sirmans 
1966:19). 

The political control of the Barbadians was challenged by the Scottish and 
English Presbyterian migrants who differed from the Anglican Goose Creek men 
in religion as well as in their place of birth and relative inexperience in 
colonial life (Sirmans 1966:40-41). Adding to this heterogeneity in back- 
ground were migrants from New England and New York, attracted to Carolina by 
its liberal land policies, and Huguenots, who were Protestant refugees mainly 
from France. Significant French Protestant migration to Carolina began in 
1679 when Rene Petit and Jacob Guerard petitioned Charles II for permission 
to lead a group of refugee families to the colony. Gue>ard led the first 
contingent of 45 families the following year, and as renewed persecutions in 
France resulted in increased migration of Protestants to England, the Hugue- 
not migration to Carolina swelled (Friedlander 1979:2). By 1700, Huguenots 
had congregated in five locations: Charles Town itself; Orange Quarter 
Creek, a tributary of the eastern branch of the Cooper River; the western 
branch of the Cooper River in what became the Parish of St. John, Berkeley; 
Goose Creek; and on the south bank of the Santee River in Craven County 
(Friedlander 1979:3). 

Eighteenth century religious wars sent new waves of Protestant refugees to 
England and her possessions. Swiss Protestants settled Purrysburg in 1732 
and New Bordeaux in 1765 (Hirsch 1928:28, 40-41). Those Huguenots who set- 
tled in the coastal parishes of the low country, the bulk of whom arrived 
prior to 1700, became indistinguishable from other colonists during the 
course of the eighteenth century. Naturalization in 1697 equalized civil 
status, except for certain restrictions on trade and office-holding that 
applied to foreign-born but not to their children who were bom in the colo- 
ny. Distinctive Calv1n1st rituals were first grafted onto the Anglican ser- 
vice and then gradually abandoned, while at least nominally conforming to the 
Church of England undermined the institutional basis of an independent Hugue- 
not interest. Although a French Protestant congregation continued to meet in 
Charles Town, its pastor in the critical decades of the early eighteenth cen- 
tury minimized the differences between his service and that of the Anglican 
Church. The rate of intermarriage between Huguenots and the rest of the popu- 
lation increased steadily until by the fourth generation, which came of age 
in the era of the American Revolution, a majority of Huguenot descendants 
married non-Huguenots. During these years, intercolonial and iritracolonial 
migration ended the pattern of partial isolation that had characterized 
Huguenot settlement at the beginning of the century (Friedlander 1979:169, 
221, 238-239, 289-290). 

During the early phases of settlement, rivers assumed great importance as 
arteries for migration. Roads, however, were quickly built, which formed a 
network linking Charles Town with the inland parishes and then with settle- 
ment deeper into the interior (Moore 1979:156). By the close of the colonial 
period, population had spread over approximately two-thirds of the present 



21 



area of the state (Petty 1943:57). The settlement of the backcountry was 
influenced by a stream of migration from the northern colonies. Although 
linked to Charles Town as a source of supply and as an outlet for agricul- 
tural surplus, this region had more a subsistence economy that produced a 
smaller surplus than a commercial /market agriculture economy. The coastal 
parishes, by contrast, invested heavily in staple crops and slaves (Petty 
1943:57). 

The Economic Development of the Low Country 

Building on the experience of Virginia and the West Indian sugar colonies, 
the proprietors, who owned the colony until 1719-1720, hoped to discover a 
staple crop whose marketing they would monopolize. The early decades, there- 
fore, saw them encouraging settlers to cultivate cotton, rice, tobacco and 
flax in addition to desirable exotic items such as olives, grapes and silk- 
worms. Adherents to the complex of ideas that constituted mercantilism, they 
believed that the colony should produce raw materials that would give the 
economically more sophisticated mother country a competitive edge in Europe's 
commerce (Friedlander 1979:75). While the Barbadians who dominated South 
Carolina in the early years had no quarrel with this essentially agrarian 
concept, they were unwilling to sacrifice their own short-term gains to the 
proprietors' long-term advantage. Specifically, trade with the Indians, 
although serving some of the settlers well, did not meet proprietary goals, 
since the proprietors had envisioned monopolizing the Indian as well as the 
white colonists' trade (Friedlander 1975:16). There was little, however, 
that the proprietors could do about the situation from London. 

Between 1670 and 1719, the Indian trade was an important factor in the Caro- 
lina economy and was critical in the early decades prior to the introduction 
of rice in the 1690s. In particular, the trade provided the capital accumu- 
lation necessary for agricultural development (Brown 1975:118). Initially, 
trade in Indian slaves was a significant part of the Indian commerce. After 
1715, exchanges of deerskins overshadowed the exchange 1n Indian slaves, who 
were rapidly lost in the multitude of black slaves (Friedlander 1975:76). 
Exports of deerskins retained their primacy until the 1740s, when the same 
individuals who had controlled the Indian trade concentrated their interests 
in rice and indigo (Brown 1975:128). 

Staple-crop agriculture was consistent with English mercantilist principles 
of empire (Haywood 1959:16). Introduced in the 1690s, rice was Carolina's 
first major staple and formed the basis for the plantation system. Overpro- 
duction in the 1730s glutted the market and was 1n part responsible for a 
decade of depression in the 1740s. Encouraged to diversify, planters began 
to cultivate Indigo. A bounty stimulated production throughout the colonial 
period and put the Carolina crop at a competitive advantage within the Bri- 
tish empire. After Independence, South Carolina indigo, which had always had 
a relatively poor reputation, lost Its advantageous position and encountered 
serious competition from indigo grown in India. Although it lingered for 20 
years, the industry finally collapsed as a result of competition, rising 
costs of production and a series of natural disasters (Winberry 
1979:249-250). 



22 



The cultivation of indigo was a labor-intensive process, requiring one hand 
per 1-4 acres on the average (C.W. 1755:202, DuBose n.d.:76, Eaton 1975:21). 
In St. Stephen's Parish, the crop was grown, cut and processed into blocks of 
dye on the plantation (DuBose n.d.:76). The land was cleared and plowed and: 

. . .after all apprehension of frost was over, the fields were 
laid off in drills about an inch deep, and from twelve to fif- 
teen inches apart from each other. In these drills the seeds, 
mixed with lime and ashes, were sown. If the season was a 
fair one, the seeds came up within ten days or a fortnight, 
and grew off rapidly. The plants were cut three or four times 
in the season, for making the dye; and during all this period 
they required nice and frequently repeated hoeing and weeding. 
When they had grown to the height of two or three feet, the 
plants were cut with a reaping hook, and carried to the macer- 
ating vat. This vat was strongly constructed of thick cypress 
planks, raised some height above the ground. When this vat, 
which was called the 'steeper,' was furnished with a -suffi- 
cient quantity of weed, clear water was poured into it, and 
the weeds were left to steep or macerate until all the color- 
ing matter was extracted from them; the fluid was then drawn 
off by means of a faucet into an adjoining vat called the 
'beater.' An axle to which were attached arms long enough 
nearly to reach the opposite sides of the vat, and each fur- 
nished with a small bucket at its end, ran lengthwise through 
the centre of this vat. Laborers would then place themselves 
upon this vat, and work the axle with handles or cranks, so as 
to cause the buckets to rise and fall alternately in the li- 
quor. This process was continued until the coloring matter 
was united in a body. . . .Lime was then applied, which assist- 
ed in the separation of the water from the indigo. The whole 
being now suffered to rest until the blue matter had settled, 
the clear water was drawn off by cocks in the sides at differ- 
ent heights, and the blue part discharged by a cock in the 
bottom into another vat. It was then strained through cloth 
bags, and spread out in shallow vessels called 'bowls,' to 
harden and dry (DuBose n.d.: 75-76). 

The dried material was cut into blocks of about one-fourth pound each and 
shipped to market in bags or boxes (DuBose n.d.: 76). 

Rice, which had supported the colonial economy, was still cultivated after 
the Revolution, but the manner of cultivation changed. In the early eight- 
eenth century, it was grown in inland swamp regions by damming ponds and 
flooding the fields in preparation for planting. In 1731, rice planting in 
South Carolina was described as follows: 

In March and April it is sown in shallow Trenches made by the 
Hough~[ and good Crops have been made without any further Cul- 
ture than dropping Seeds on the bare Ground and covering it 
with Earth, or in little Holes made to receive it without 
further Management. It agrees best with a rich and moist 
Soil, which is usually two Feet under Water, at least two 
Months 1n the Year. It requires several Weedings till it is 



23 



upward of two Feet high, not only with a Hough, but with the 
Assistance of Fingers. About the middle of September it is 
cut down and housed, or made into Stacks till it is thresh' d, 
with Flails, or trod out by Horses or Cattle; then to get off 
the outer Coast or Husk, they use a Hand-Mill, yet there 
remains an inner Film which clouds the Brightness of the 
Grain, to get off which it is beat in large wooden Mortars, 
and Pestles of the same, by Negro Slaves, which is very 
laborious and tedious (As quoted in Rasmussen 1975:160-161). 

Gideon Dupont is credited with initiating the change to tidal culture in 
1783. There is evidence, however, that methods taking advantage of the ef- 
fect of the tides on rivers and swamps in order to flood the rice fields were 
in use prior to the 1780s. Flooding the fields killed the weeds but not the 
rice plants, thus obviating the need for repeated hoeing and lessening the 
intensity of labor required to cultivate rice. Rice in nineteenth century 
South Carolina, according to Frederick Law Olmstead in 1853: 

. . .continues to be cultivated extensively on the coast of 
Georgia and the Carol 1nas. . .only because there are unusual 
facilities there for forming plantations in which, while the 
soil is exceedingly rich and easily tilled, and the climate 
favorable, the ground may be covered at will with water, until 
nearly all other plants are killed, so as to save much of the 
labor which would otherwise be necessary in the cultivation of 
the crop; and which may as readily be drained, when the 
requirements of the rice Itself make 1t desirable (as quoted 
in Rasmussen 1975:815-816). 

Clarence Ver Steeg (1975:120) has argued that naval stores were the basis for 
the development of plantation culture and slavery in the early eighteenth 
century in South Carolina. Mr. George Terry, Curator, Historical 
Collections, McKlssick Museum, has graciously supplied parts of the draft of 
his yet unpublished study of the Parish of St. John, Berkeley 1n the eight- 
eenth century. In this study, he disputes Ver Steeg's finding, arguing that 
his conclusions rest on faulty reading of rice export statistics, although he 
would not deny that production of naval stores was a significant feature of 
the colonial economy (Terry 1981:11:26-31). Forest industries, in both 
Georgia and South Carolina, "figured Importantly as a wintertime activity 
that created year-round labor for the slaves and servants, provided a source 
of supplementary Income, and cleared land for agricultural expansion" (Hern- 
don 1979:131-132). Few planters, especially those in newly settled parishes, 
were so well -fixed that they could afford to allow their slaves to be idle 
and to ignore the profits of their woodlands. 

St. Stephen's Parish 

The Indian trade stimulated endemic warfare among the tribes, and constant 
abuses by the whites resulted in Indian wars against the Europeans (Fried- 
lander 1975:39). The Yamassee War of 1715-1716 devastated the colony, and 



24 



subsequent mishandling by the Proprietors provided the imperial government 
the opportunity to obtain control of the colony in 1719-1720 (Sirmans 
1966:129). At this time, North Carolina and South Carolina were formally 
separated; other agencies of local government, i.e., the parish and the 
assembly, were left intact. 

The primary unit of local government was the parish (Figure 3). In 1682, the 
vicinity of the Ashley and Cooper Rivers was divided into three counties: 
Berkeley, Craven and Colleton (Archives News 1968(69 ) : 1 54) . These divisions, 
however, did not assume administrative importance. Despite official tolera- 
tion of non-Anglicans, the ten Anglican parishes established in 1708 were 
units of social, civil, political and religious importance (Hannum 1970:40). 
The vestrymen and churchwardens were men of local influence. They hired the 
minister, supervised maintenance of roads and schools, and assessed taxes for 
the care of the poor (Boucher 1948:1). The parish also served as the elec- 
tion district for representation in the Commons House of Assembly, the colony 
representative body after 1716 (Hannum 1970:55). As the colony grew, ad- 
ditional parishes were surveyed, responding to a diffusion of population into 
the interior and to shifts in population density. Thus, the area known in 
the early eighteenth century as English Santee, a part of the Parish of St. 
James, Santee, was incorporated as the Parish of St. Stephen in 1754 as a 
result of movement into the area during the preceding decades (Misenhelter 
1977:1). 

For most of the eighteenth century, swamplands and river lands were desired 
as avenues to wealth. The population in the coastal parishes spread out 
along a network of swamps that intertwined among the rivers. People in St. 
John's, Berkeley, and St. James, Santee, followed the swamps that spread 
along the Cooper and Santee Rivers (e.g., Hell Hole Swamp, Ferguson's Swamp, 
Wiskinboo Swamp, Half Way Swamp, Santee River Swamp, Fair Forest Swamp), 
gradually filling in the area called English Santee, subsequently called St. 
Stephen's (Orvin 1973:5). With swamplands suited for rice and higher land 
for indigo, the parish prospered (Figure 4). 

The Parish of St. Stephen was very like the two neighboring parishes that had 
influenced its settlement: St. James, Santee, and St. John's, Berkeley. Des- 
cendants of Huguenot migrants were prominent in parish government in the lat- 
ter eighteenth century, and St. Stephen's reliance upon plantation staples 
was very like the economy of the two adjoining parishes. The Huguenots who 
resided in St. Stephen's were the grandsons and great-grandsons of men who 
had settled in Santee; St. John's, Berkeley; and Orange Quarter. Like other 
members of the colonial elite, leading planters dominated parish affairs, and 
in this parish, many leading planters were descendants of Huguenot migrants. 
In St. John's, Berkeley, the early eighteenth century witnessed the emergence 
of kinship networks within the parish, and in St. Stephen's, a similar pheno- 
menon developed. An analysis of the seven member vestry between 1754 and 
1770 indicates that in 13 of the 17 years, a majority of the members of the 
vestry were Huguenots. In 14 of the 17 terms for which the membership is 
known, more than half of the Huguenot members were kindred (Fried! ander 
1979:285-286). 



25 





26 



3JELE£LB£LNC 



_EP=- 



_/.,»*-■«*/ * tw* +f Art »/**»***—*/* . |i 



^r.r^fm ***** m*a, 

**~,->*A A /*•#*_ _ ._ ______ ".' "J 

J~-~-,*, . _ 

C^v 'fA*S _ _____ 



^ 



>H 



fro*** <SiV-^'w4 Cr*»A A* /-*—~*r* ±3*-*&fm 



: :-6_t 



3: 



a Map 

of the 

PARISH 

of 

S T STEPHEN, 

in 

CRAVEN COUNTY 

Exhibiting aVicw oflhe fevernl Places 
Practicable far making a 

Navigahlk Canal 

between 

San tee and Co oper R ivers 

from an Actus 1 Survey 

7?y / lENHYA tO UZ OJK/u $ 

17 7.1 

A SCALE OFM1U2SAND CHAINS 

6g£ Miles to a Degree 



^Wfe \_«- 



is3 

:t_i__-_4-: 









7 



*v--..M 



_/ -53£?rfe_ 






,->>H 



US 



B 









.^ 






~* » ' 



"lit " 



-* '-.vl 



;^.>' • --A r-V j'Jfi FIGURE 4 
^^-.'^rj^fit^tja Mouzon Map, 1771 
■-.".4 ■ {■ .-%:-••.:-•■*;» •. i^t'^.-S*. 




26 



*&> "" " \L I '■ 1. v^^< ,4i^\. --""^ 



mm 




-fiFFFffFMf^- 









fie?)' 



»*""" .""'nKii" ■ 



a Map 

if the 

PARISH 

of 

S T STEPHEN, 



CRAVEN COUNTY 

Exhibiting aA/icw oflhc feveruj Places 
Pracltablt for making a 

M-wiGAHiJK Canal 

between 

San tee and Co ofer R ivers 

fnini in AcIiihI Survey 

■Byl lKNHyMOL'ZONjur, 

IV 71 

ASC. ML ufM!L*:SyiND CHAINS 

6gJ Miles to a Degree 



fe&---^ 



■>£3sH- 



i},;$8iitftf- 



f ^t 



- t~"5 



:KSi 



f^VALLEVl^vv_ / 



./*' 



^m. 









i*v. •Jiiji*-': 



*S 



5- V 



!?-/*?• 






- %sSSfc 












>%- 



• rfrr.'p'fe 






-.-: ; £ 3 £|$ FIGURE 4 
o'^rS^n^M Mouzon Map, 1771 



27 



Although rice and indigo supported the colonial South Carolina plantation 
system, the prevalence of this model has been overestimated. George Terry's 
research on the Parish of St. John, Berkeley, shows that particularly in the 
early stages (i.e., the late seventeenth and early eighteenth centuries), the 
economic organization in the parish "defies any simple description" (Terry 
1981:11:37). Individual planters experimented with everything "from cotton 
to hemp" and planters had sizable investments in corn, cattle and naval 
stores in addition to rice (Terry 1981:11:35-36). In the interests of 
protecting the integrity of their families, Terry argues, planters in the 
early eighteenth century acquired land outside the parish limits on which 
they envisioned settling their children, moving, he states, generally in the 
direction of the Santee River (Terry 1981:11:38). Terry's preliminary 
conclusions are consistent with other detailed studies of the low country in 
the region involving the Cooper and the Santee Rivers, and suggest clearly 
that the kinship network evident in St. Stephen's Parish was the natural 
evolution from patterns evident in neighboring St. John's, Berkeley. The 
geographical direction was influenced by the needs of South Carolina's 
plantation economy and resources of the forest. 

The War for Independence substantially altered the economics of St. Ste- 
phen's. Although planters continued to work with indigo, they began to make 
the transition to cotton culture, facilitated in the 1790s by the invention 
of the cotton gin and the completion of the Santee Canal. Neglect, slave 
confiscations and dislocations caused by the war had a deleterious effect on 
the economy of the parish. At the beginning of the Revolution, Peter Gail- 
liard of St. Stephen's Parish inherited a good indigo plantation in the 
parish; ten years after the war, he could not make ends meet. Around 1790, 
Samuel Cordes abandoned Mil ford, for which he had paid 6000 guineas before 
the Revolution. Gailliard purchased The Rocks, a plantation in Upper St. 
John's, Berkeley, above St. Stephen's, where he successfully cultivated 
cotton after 1796. By 1800, he was out of debt, and when he died he left 
each of his five sons a plantation and each of his three daughters a town- 
house, in addition to the comfortable fortune which had seen him through his 
old age in comfort (Stoney 1938:39). Building the Santee River Canal had 
provided limited financial relief for Santee River planters whose slaves were 
employed by the project between 1792 and 1800. By the opening of the canal, 
enough acreage was in cotton to make the crop one of the significant items 
carried along it (Orvin 1973:148-149). 

Although cotton proved a godsend for Upper St. John's, Berkeley, and for the 
upper part of St. Stephen's Parish, particularly the vicinity of Pineville, 
which was founded in 1794, the parish as a whole suffered from the war and 
from a rapid succession of freshets in the late eighteenth and early nine- 
teenth centuries. In 1826, Robert Mills commented: 

At present there are many waste old fields, both high 
lands, and river swamp, which, thirty years ago, were in the 
highest state of cultivation, producing luxuriant crops of 
corn, indigo, and rice. This melancholy reverse is the effect 
of freshets, no measures being taken to bank in the river 
lands from the flood. 



28 



These lands are uncommonly fertile, and were successfully 
cultivated till the year 1784. From that year to 1795, very 
little was raised near the Santee. Many of the planters, 
discouraged by a rapid succession of freshets, abandoned the 
plantations subject to their influence. Since 1796, the 
freshets have diminished in frequency and height; and the 
planters have recommenced there the culture of corn, rice, and 
cotton (Mills 1826:482). 

Plantation records describing the holdings of Thomas Porcher and Thomas 
Walter Peyre do show these men holding several tracts that had been eight- 
eenth century plantations consolidated into larger units. Known by their 
eighteenth century names, they were planted in corn, cotton, peas and oats 
(Thomas Porcher, Diary, 1822-1824, SCHS; Thomas Walter Peyre, Journal, 
1834-1850, SCHS). 

Mill's optimism was, however, premature. The Santee River canal became ex- 
pensive to maintain, although the construction of the North Eastern Rail Road 
compensated for the loss of the canal as an outlet (Orvin 1973:153-154). 
Pineville suffered a series of epidemics between 1833 and 1836, which result- 
ed in temporary depopulation; its academy did not reopen until 1842 (Orvin 
1973:158-159). Finally, cotton cultivated in coastal Carolina was oversha- 
dowed by the tremendous profitability of the Piedmont and then the Gulf 
States, particularly Alabama, Mississippi and eastern Texas. 

South Carolina and Cotton Production, 1800-1860 

"Upland" (gray and green seed) cotton was introduced as early as 1733 and 
spread rapidly around 1800 as a result of the demand created by the indus- 
trial revolution, the exploitation of black labor and the invention of the 
cotton gin. Eli Whitney's gin, however, was more significant in the spread 
of green seed cotton, and hence to the development of the Piedmont and the 
Gulf states, than to changes in the coastal regions, since black and gray 
seed cotton, the varieties grown in the east, were already effectively 
cleaned through a gin that antedated Whitney's machine (Brasington 
1977:14-15). In the first 20 years of the nineteenth century, South Carolina 
"began to enjoy something of a monopoly in cotton production" so that by 
1825, "cotton had pushed almost everything except articles for home consump- 
tion into the background" (Whartenby 1977:21-22). The price of cotton fell 
in 1819 and remained low through the 1820s. During these less prosperous 
years, Alabama and Mississippi entered the Union, and the cotton planters in 
these states "could make profits at prices which would bankrupt many of the 
South Carolina planters" (Whartenby 1977:23). 

Between 1801 and 1811, the production of cotton in the inland coastal par- 
ishes more than doubled. Between 1811 and 1821, production increased by 24 
percent, but the national statistic, affected by the production in the Pied- 
mont and settled parts of Alabama and Mississippi, showed an increase of 121 
percent. In the years from 1821 to 1839, the production of the inland 
coastal parishes decreased by 25 percent. South Carolina's total production 



29 



also fell by 10 percent in this period, although national production in- 
creased by 123 percent. In the 1840s, South Carolina's production improved, 
growing by 87 percent; the coastal parishes increased production by 210 per- 
cent in this decade. National production increased only 25 percent in the 
1840s. The forties were a decade of depressed prices, but the fifties were 
flush. In the decade preceding the Civil War, national production again went 
up by 140 percent; the inland parishes increased their yields by 60 percent, 
and the state's production went up only 18 percent (Brasington 1977:20-21). 

One of the ways in which the inner coastal plain retained its profitability 
in cotton was through the early use of fertilizers and the consolidation of 
farms into larger plantations (Brasington 1977:18). This resulted in the 
displacement of small farmers by larger property holders. The migration from 
the Seaboard to the Gulf states was dominated by small farmers who did not 
own slaves; slaveholders followed. The small farmer, James 0. Foust con- 
cludes, "does not appear to have been pushed out of the better lands in the 
Southwest. In fact, more 'pushing out' appears to have occurred in the older 
southeastern regions" (Foust 1975:171). Thus, although the inland coastal re- 
gion continued to engage profitably in cotton agriculture in the six decades 
prior to the Civil War, it declined relative to the rest of the South and its 
limited prosperity took place at the expense of smaller operations. Sadly 
and with some bitterness, James Hammond told the Agricultural Society of 
South Carolina in May 1842: 

. . .the cotton growers of South Carolina need not look abroad 
for competition. It is much nearer home. It is our own kith 
and kin. . .that have levelled the gigantic forests of the 
south and southwest, and furrowed the rich bottoms through 
which pour the tributaries of the Gulf of Mexico, from the 
Suwanee to the Sabine, and that have but recently rescued from 
a slothful race the fertile empire stretching beyond the 
Sabine to the Rio Grande -- who are destined at no distant 
day, to supply the foreign markets of the world with this 
inestimable staple. . .so soon as the check on consumption 
shall place in strict competition all the cotton growers of 
the world, and reduce prices to their lowerst point, the 
cultivation of this staple must be confined almost entirely to 
these fertile regions (as quoted in Rasmussen 1975:711-712). 

In the meantime, observers of St. Stephen's and the older settlements along 
the lower Santee in the late antebellum period saw decayed grandeur. "He who 
travels in winter from the bank of the Santee Canal" [i.e., down the Santee 
through St. Stephen's], wrote Fredrick Porcher, "will find himself in an al- 
most uninterrupted forest of pine. On his left lie .the mysterious depths of 
the Santee Swamp, whose soil once teeming with the rewards of industry, is 
now abandoned to the hand of nature. ..." (Porcher 1852:93). Continuing 
along the river, Porcher described past and present evidence of making tar, 
and arriving at the parish church, he commented: 



30 



The church tells a story of former grandeur and of present 
desolation; though not large, it indicates a respectable 
congregation; it is finished with neatness, with some 
pretensions even to elegance, and the beholder involuntarily 
mourns over the ruin to which it is doomed (Porcher 1852:95). 

Dr. Samuel DuBose also pondered the question of the decay of St. Stephen's in 
the 1850s, remembering the parish in his childhood as having been "the garden 
spot of South Carolina" (DuBose n.d.:37). Reflecting on first the church and 
then the vacant plantation houses, he commented: 

Silence is becoming there [in the graveyard]; it is what we 
naturally expect. But here, in the abiding-place of men, 
where was once the din of busy life, we have now the silence 
of death, and more than its gloom. For these walls were meant 
for the living, but now no living soul dwells within them 
(DuBose n.d.:85). 

He exaggerated the extent of depopulation since federal censuses continued to 
enumerate people in the parish (Table 1 and Table 2). The population was for 
the most part concentrated in the upper end of the parish, particularly in 
Pineville, which became a retreat for planters and their families, populated 
year-round by a clergyman, doctor, storekeeper and an assortment of widows 
and spinsters (Brewster 1947:44). 

Slavery in South Carolina 

The most influential migrants in the 1670s were a group of Barbadians who 
migrated to the American mainland because of a growing scarcity of land in 
the islands. With them, they brought familiarity with plantation management, 
staple-crop agriculture, and slavery. Eleven percent of the households that 
arrived in Charles Town between 1670 and 1680 included slaves (Friedlander 
1979:100-102). Early settlers in the colony, however, evidently expected to 
enslave the native population, and the reports back to the proprietors in 
England were soon filled with accounts of exploiting and enslaving the Indian 
tribes. The Westo War of 1684 effectively decimated the most powerful of the 
coastal tribes in the area, and the Yemassee War of 1715-1716 broke the back 
of this powerful group, which had acted as a buffer between the European 
population along the coast and the more powerful tribes of the interior. 

The Indian trade involved traffic in slaves as well as skins and furs. 
Slaves were generally captives taken as a result of inter-tribal warfare, 
which the traders did their best to incite. Indian slaves were re-shipped to 
other colonies on the mainland or in the West Indies as well as kept in South 
Carolina. Because of the ease with which they could escape and the threat 
they represented, the colonial government tended to encourage the exportation 
of Indian slaves. As measured by the proportion of the slave population, the 
Indian presence peaked by 1710 and was quickly overshadowed by the enormous 
influx of Africans after 1710 (Table 3). Recent estimates developed by 
Philip Morgan (1977:284) indicate that the average number of African slaves 



31 



<u — > 

£■ — 

CU 3 



GO 



c o 
,- to 
— 00 
o — 



o 

1- CU 
•a *j 



•S 3 

O 



_, 


r^ 


O 


LO 


f»> 


*r 


m 


- 


-" 


— 


— ' 


— ' 


"■" 




r*» 


Cn 


CM 


*r 


O 


»r 


ON 


Cn 


-o 

CO 


LO 

CO 


CM 
CO 


LO 




CO 


LO 

f44 


o 
lo 
cm 


CM 

CO 
CM 


CM 


cn 

-O 
O 
CM 

1 — 1 


CO 


CM 

cn 


LO 

LO 

CM 


m 


,_ 


U3 


_ 





LO 


_ 


09 


m 


«r 


- 


LO 
CM 


Cf> 


CM 


CM 

cm 


a 

m 
cn 


r— 1 
CM 

cn 


CM 


CM 
O 
LO 


CO 


CTi 

CO 

LO 


O 


o 


O 


O 


O 


O 





o 
o 


o 
o 


O 

o 


O 



O 
O 


O 

O 







m 

p*- 
CM 


CM 

lo 

CM 


« 

41 

un 

CM 


m 
O 

LO 

CM 


LO 
CM 


ro 

LO 

*»• 

CM 


IT) 

ao 

CM 


LO 


-o 


en 


- 


CM 


CM 




O 


CM 


m 


*T 


r»> 


O 


O 




cn 




m 


LO 


LO 
LO 


on 

LO 


cj> 

o 

o 


m 
en 


lo 
cr> 


ao 

LO 

CM 


cn 


00 

cn 
O 

CM 

cn 


on 

m 
cn 
m 


o 


CO 


r-* 


LO 


r*. 


O 


O 


LO 


V0 

LO 


CM 


CO 


lO 


LO 


JJJ 


a* 

p»* 


LO 
CM 


o 

tft 


LO 
CM 


ao 




LO 

cn 


m 

LO 
LO 


«r 


43 

cn 


CO 

CM 


CM 


LO 

LO 
CM 


LO 
CM 


CM 


O 


O 


O 


O 


O 


O 


O 


O 

© 


o 
o 


o 

o 


O 
O 


O 
O 


O 
O 


O 
O 


O 


cn 

LO 


LO 




LO 

ao 


CO 

ON 
en 


O 

LO 


cn 

CM 


lo 

m 


LO 


CM 
O 
LO 


co 

LO 


cn 

LO 


CO 

LO 


o 

p>^ 


O 

o 

CO 


O 
CO 


O 
CM 
CO 


O 
cn 
ao 




CO 




LO 

co 



0) 






jr 




a. 


a. 




Q. 


cu 






w 






LA 




O 
O 


4-> 




<a 


LO 


CO 





LO C-4 



• o 
1/1 

L. 
CU 



e e a o 



cu ■»-> 

o 
• «o 

O cn 

lo r» 
ao <— 



3 3 u» kw 



= te x =s 



o 

00 



LO 


141 


CT> 3 


1 LTt 


*r c 


cn CU 


u 


CL.C 


o.-- 


X 




— (/> 





m cu 


CO £ 


<— w 


*- 


l/t 


O 


T 




Ul 


E 


e 


3 


a> 




«_> 


T3 




C 


JB 


<u 




r^ 




O 


U- 


U 



32 



TABLE 2. Population of South Carolina, 
Berkeley County and St. Stephen, 1890-1970 

South Carolina Berkeley County St. Stephen 

Year Total % Increase Total % Increase Total % Increase 



1890 


1,151,149 


15.6* 


55,428 


— 


230 


— 


1900 


1,340,316 


16.4 


30,454 + 


-45.0 


256 


11.3 


1910 


1,515,400 


13.1 


23,487 + 


-22.9 


408 


59.4 


1920 


1,683,724 


11.1 


22,558 


- 3.9 


[546]** 


33.8 


1930 


1,738,765 


3.3 


22,236 


- 1.4 


[616] 


12.8 


1940 


1,899,804 


9.3 


27,128 


22.0 


1185 


92.4 


1950 


2,117,027 


11.4 


30,251 


11.5 


1341 


13.2 


1960 


2,382,594 


12.5 


38,196 


26.3 


1462 


9.0 


1970 


2,590,516 


8.7 


56,199 


47.1 


1506 


3.0 



* Population of state in 1880 was 995,577; Berkeley County was organized 
in 1882. 
** Brackets indicate estimates based on trend-line analyses. 
+ Portions of Berkeley County were annexed to Charleston, Orangeburgh and 
Dorchester Counties between 1890 and 1910. 

Sources: U.S. Census of Population: 1900 , I, p. 555. 

U.S. Census of Population: l"9T0~ , III, p. 643. 
U.S. Census of Population: 19"?0~ , III, p. 929. 
U.S. Census of Population: 1910" , III, pp. 784, 806 
U.S. Census of Population: 19W , II, pp. 376, 420, 346. 
U.S. Census of Population: 1930" , II, Pt. 40, pp. 40-9, 

40-74, 40-65, 40-26. 
U.S. Census of Population: 1960 , I, Pt. 42, pp. 42-9, 

42-45, 42-74, 42-22. 
U.S. Census of Population: 1970 , I, Pt. 42, pp. 42-7, 

42-12, 42-41, 42-89. 



33 



TABLE 3. South Carolina Slave Population, 1703-1724 
Total Black % of Total Indian 



% of Total 



Year 



1703 


3350 


3000 


89.6 


350 


10.4 


1708 


5500 


4100 


74.5 


1400 


25.5 


1710 


[6740] 


5500 


81.6 


[1240] 


19.4 


1715 


12350 


10500 


85.0 


1850 


15.0 


1720 


13888 


11828 


85.2 


2060 


14.8 


1724 


[16554] 


[14454] 


87.3 


2100 


12.7 



♦Brackets indicate estimates based on trend line analyses 



Sources: Records in the British Public Record Office V:203 
Clowse 1973:252 
Greene and Harrington 1932:172-73 



34 



imported into South Carolina per year in the period 1706-1709 more than 
tripled in the period 1710-1714. The average number of imported slaves 
continued to grow, reaching a peak of 2370 in the period 1735-1739. 

By 1710, slaves outnumbered free people in South Carolina. Peter Wood 
(1974:274) has argued that after 1720, living in the presence of a black 
majority engendered a white mind-set anxious about incipient rebellion and 
simultaneously defensive in matters relating to the question of loyalty of 
slaves to their masters. Research by George Terry (1981) in St. John's, 
Berkeley, and Philip Morgan (1977) on the evolution of plantation slavery in 
Virginia and South Carolina in the eighteenth century has detailed with some 
precision the extent of the black majority and the significance that this had 
upon the nature of South Carolina society in the colonial period. 

Morgan (1977:1) found that in the 1730s slaves tended to become concentrated 
on larger plantations; half of the colony's slaves lived on plantations with 
20 to 50 slaves. Although a small increase in the size of the slave popula- 
tion was apparently due to natural causes in the early eighteenth century, 
the enormous growth due to importation between 1715 and 1739 and again in the 
1750s more than accounted for the increase in the slave population. Popula- 
tion growth due to natural increase was not again evident until the 1760s 
(Morgan 1977:288). By 1790, in the Santee region, which included St. Ste- 
phen's Parish, three-fourths of the slaves were resident on Targe plantations 
(Morgan 1977::!). Dealing more specifically with society in the Parish of 
St. John, Berkeley, Terry (1981:111:1) has found that a "scarcity of white 
population" was "one of the most distinctive features" of the parish. He 
believes that the average ratio of blacks to whites was 15:1 during the 
eighteenth century (Terry 1981: IV: 15). At the close of the eighteenth cen- 
tury, Morgan (1977:7) argues that on the large plantations, which housed 
three-fourths of the parish's slaves, the ratio of slaves to whites was 27:1. 

The numbers themselves are less important than their significance relative to 
black culture. The exaggerated imbalance meant that slaves were in less fre- 
quent contact with whites and hence less likely to be in a position to graft 
colonial, white customs onto their own practices. Although Terry (1981:IY:2) 
estimates that the white population in the parish never exceeded 600 during 
the eighteenth century, the black population, in contrast, grew to nearly 
4000 in the same period. Blacks, particularly in the outlying plantations 
owned by families that held several plantations, were more likely to have 
been left on their own. Although plantations wholly composed of slaves were 
the exception rather than the rule, Terry (1981: IV: 15-18) uncovered suffi- 
cient evidence to demonstrate limited black autonomy within the plantations 
in St. John's, Berkeley, and a surprising amount of individual mobility on a 
regular basis within the parish, into the adjoining parishes of St. Thomas 
and St. Denis, and along the Cooper River. 

Census data for St. Stephen's Parish begins in 1790, although that census was 
not a definitive enumeration. Census data collection improved with time, and 
the censuses of 1840 and 1850, particularly the enumerators' manuscript re- 
ports, are more reliable. Table 2 summarizes census data for South Carolina 
and St. Stephen's Parish from 1790 to 1850, with emphasis on the relationship 
between the white and black populations. 



35 



On a state-wide basis, the balance between blacks and whites had reversed 
from its colonial trend in 1790. Reflecting, however, the extension of plan- 
tation slavery into the Piedmont as a result of cotton cultivation, the ratio 
of blacks to whites reverted toward its colonial arrangement in the first 
decades of the nineteenth century. By no means, though, did antebellum South 
Carolina's population resemble the extreme racial imbalance of the early 
eighteenth century colony. Because St. Stephen's had been an eighteenth cen- 
tury plantation parish, with an extremely high concentration of large planta- 
tions, the ratio of blacks to whites was higher in 1790 than for the state as 
a whole. Over the course of the first five decades of the nineteenth cen- 
tury, this ratio tended to approach the state statistics, but reflecting the 
investment in plantation agriculture over smaller farms, the ratio of blacks 
to whites in St. Stephen's was consistently higher than the state's ratio. 

Table 1 1 1—4 suggests at least one reason for the decrease in the racial imbal- 
ance in the parish. Clearly, except for the decade of the 1830s, white popu- 
lation grew more rapidly than black. Through the 1830s, there was a consis- 
tent decline in the black population 1n the parish, which was probably the 
result of the movement of plantation agriculture into the Piedmont. White 
depopulation of the parish, characteristic of the 1830s and consistent with 
contemporary remarks about the state of the parish (see preceding section), 
reversed itself in the 1840s. By the 1850s, the populations of both groups 
were growing. Because both groups grew, the imbalance stabilized at between 
4:1 and 3:1 and neyer became as extreme as it had been in the previous cen- 
tury. Consequently, contact between whites and blacks in the first five 
decades of the nineteenth century in St. Stephen's can be assumed to have 
been more frequent, and the demographic conditions conducive to black 
autonomy in the colonial period weakened in the nineteenth century. 

Ownership of Slaves 

Scholars have estimated that at the close of the colonial period, between 
one-half and three-fourths of eastern South Carolina's white population held 
slaves (Main 1966:65; Jones 1980:13). The most recent study of wealth dis- 
tribution at the conclusion of the colonial period by Alice Jones (1980:13) 
finds, however, that in 1774 "a few of them held large numbers of slaves". 
Philip Morgan (1977:7) has argued that in 1790 in parishes consisting almost 
entirely of large plantations, such as St. Paul's and St. John's, Colleton, 
only 5 percent of the households did not report any slaves and "only in the 
large parishes of Prince George's and Prince Frederick's to the north were 
about one-half of the households without slaves". The two northern parishes 
were toward the colonial frontier, and real estate was more equitably 
distributed than in the coastal plantation parishes. In the latter 
eighteenth century in South Carolina, therefore, mere ownership of a slave 
was within the economic grasp of at least half of the population. Although 
owning slaves was fairly widespread within the white population, owning large 
numbers of slaves was restricted to an elite. Slaves constituted about 90 
percent of the total population of St. Stephen's Parish 1n 1790, and 
therefore, when historians discuss the major colonial slaveholders, meaning 
that upper 30 percent of slaveholders (or 3 percent of the total population) 
who owned 26 or more slaves 1n 1774 (Jones 1980:119), they are truly 
describing the elite of the elite. 



36 



TABLE 4. Population Growth, St. Stephen's Parish 

1790-1850 



Year White % Increase Black % Increase 



1790 


227 


— 


2506 


— 


1800 


330 


+45.0 


2182 


-12.9 


1810 


[372]* 


[+12.7] 


[2174] 


[-.37] 


1820 


[427] 


[+14.7] 


[2063] 


[-5.1] 


1830 


602 


+41.0 


1814 


-12.1 


1840 


481 


-20.0 


1972 


+8.7 


1850 


689 


+43.2 


2165 


+9.9 



* Brackets indicate estimates based on trend-line analysis. 



Sources: Ms. Census, 1840, Charleston District, St. Stephen's Parish 
Ms. Census, 1850, Charleston District, St. Stephen's Parish 
U.S. Census: 1790 South Carolina, p. 8. 
Return of the Whole Number of Persons (1800), pp 2-3. 
Census for 1820. 
Fifth Census (1830), pp. 94-95. 
CompendiunTof the Sixth Census (1840), p. 367. 



37 



Morgan (1977:6-7) has shown that the South Carolina slave population tended 
to become concentrated on large plantations during the colonial period. 
Since slaves continued to be imported throughout the colonial period, this 
did not affect the distribution of slave ownership within the white popu- 
lation. After 1808, the transatlantic slave trade closed, and growth of the 
black population was the product of changes in life expectancy and fertility. 
A more limited supply might be expected to accelerate the concentration of 
slaves in the hands of fewer masters so that ownership of slaves would be 
less widespread in the white population. 

The scholarly literature shows that slave ownership did become more re- 
stricted, but also demonstrates variation within the region of individual 
states. Based on the printed 1860 census, Kenneth Stampp (1956:31) found 
that nearly three-fourths of the white population did not own any slaves at 
all. This differed from state to state. In South Carolina and Mississippi, 
about one-half of the families reported slaves. In Georgia, two-fifths of 
the households owned slaves, and the proportion declined to one-thirtieth in 
Delaware. Large si avehol dings were more numerous in the Lower South than in 
the Upper South, and "concentrations of the southern plantation could be 
found in the sugar parishes of Louisiana, in the Yazoo Basin, and around 
Natchez in Mississippi, in the Black Belt of Alabama, and in the rice swamps 
and sea islands of South Carolina and Georgia" (Stampp 1956:31). Almost 50 
percent of the slaveholders held less than five slaves, 72 percent owned less 
than ten, and 88 percent had less than 20 slaves, the benchmark in the ante- 
bellum decades for large holdings. The "typical" planter, according to 
Stampp, worked between 20 and 50 slaves, and the aristocracy consisted of 
some 10,000 families who owned between 20 and 50 slaves. "The extremely 
wealthy families who owned more than a hundred slaves number less than 3000 
in a total of 1,516,000 free families, a tiny fraction of the southern popu- 
lation (Stampp 1956:30-31)' 



M 



The plantation culture that had been geographically confined for the greater 
part of the colonial period had thus diffused across the South during the 
years from 1790 to 1860. Society was not static, and Gavin Wright's studies, 
based on the manuscript censuses for 1850 and 1860, show that the 1850s had 
seen some shifts in slave ownership. Wright (1970:81) found that the social 
segment just below the very top, "the second, third, and sometime fourth 
deciles" gained at the expense of both the richest and poorest members of 
Southern society in the decade preceding the Civil War. He concludes that 
"the small farmer was not so much being squeezed off his land as losing his 
share of the slaves. These two factors — the rise of slave prices and the 
shift of slaveownership away from small farmers toward middle-class farmers 
-- resulted in an increase in the concentration of wealth for the cotton 
South as a whole". 

Analysis of the manuscript census for the Parish of St. Stephen for 1840, 
1850, and 1860 exemplifies the generalizations outlined by both Wright and 
Stampp (Table 5, a-c). Wright s argument implies that the distribution 
Stampp summarized evolved from a broader distribution of slave ownership in 
the earlier decades. Between 1840 and 1860, the percentage of white house- 
holds in St. Stephen's reporting slaves did decline, although the increase in 
1850 probably reflects the parish's relative prosperity in that decade, de- 
spite economic decline in most of the cotton South in the 1840s. 



TABLE 5a. Frequency Distribution, Slave Ownership, 
St. Stephen's Parish, 1840 

Interval Number PerceTvt Cumulative % of Slaveholding Cumulative 



38 






57 


1-4 





5-9 


5 


10-19 


11 


20-50 


9 


51-100 


6 


100+ 


6 



60.6 


60.6 





60.6 


5.3 


65.9 


11.7 


77.6 


9.6 


87.2 


6.4 


93.6 


6.4 


100.0 



13.5 
29.7 
24.3 
16.2 
16.2 



13.5 
43.2 
67.5 
83.7 
99.9* 



Total 37 

Range, 0-260 

*Error due to rounding 

Source: Ms. U.S. Census: 1840, South Carolina, Vol. 2 



Mean Slaveholding = 53.4 



TABLE 5b. Frequency Distribution, Slave Ownership, 
St. Stephen's Parish, 1850 



Interval Number Percent Cumulative % of Slaveholding Cumulative" 






74 


1-4 


17 


5-9 


16 


10-19 


11 


20-50 


13 


51-100 


2 


100+ 


7 



52.9 
12.1 
11.4 
7.9 
9.3 
1.4 
5.0 



52.9 
65.0 
76.4 
84.3 
93.6 
95.0 
100.0 



25.8 
24.2 
16.7 
19.7 
3.0 
10.6 



25.8 
50.0 
66.7 
86.4 
89.4 
100.0 



Total ~~55 

Range, 0-369 

Source: Ms. U.S. Census; 

TABLE 5c. 



Mean Slaveholding = 44.9 
1850, South Carolina, Vol. 2 

Frequency Distribution, Slave Ownership, 
St. Stephen's Parish, I860 



Interval" Number Percent Cumulative % of Slaveholding Cumulative 






147 


1-4 


14 


5-9 


13 


10-19 


12 


20-50 


10 


51-100 


3 


100+ 


5 



72.1 


72.0 


6.9 


79.9 


6.4 


85.3 


5.9 


91.2 


4.9 


96.1 


1.5 


97.6 


2.4 


100.0 



24.6 
22.8 
21.0 
17.5 
5.3 
8.8 



24.6 
47.4 
68.4 
85.9 
91.2 
100.0 



Total 57 

Range, 0-243 Mean Slaveholding = 29.1 

Source: U.S. Census, 1860, South Carolina, Vol. 2 



39 



By 1850, moreover, the number of small si avehol dings had increased, offset- 
ting the increase in the absolute number of plantations housing more than 100 
slaves. St. Stephen's was on the edge of the very rich plantation coastal 
area, and like the coastal parishes, it consisted of a disproportionately 
large number of plantations housing over 100 slaves. The "typical" planter 
in St. Stephen's worked somewhat more slaves than the "typical" Southern 
planter Stampp described, although the mean size of slaveholding consistently 
declined between 1840 and 1860. Finally, the distribution of slave ownership 
within the slave holders seemed to remain fairly stable after 1840. The 
white and black population remained largely stable until 1850, and modest 
growth in the parish's total population evidently included the migration into 
the parish of non-sl avehol ding farmers and small planters, particularly be- 
tween 1850 and 1860. Between 1850 and 1860, moreover, the absolute number of 
slaveholding households dropped although the parish's total number of house- 
holds increased. The distribution of slaveholding households on the eve of 
the Civil War suggests that very large plantations had been curtailed as had 
the very small, a pattern which is consistent with Wright's conclusions. 
Slaveholders throughout the antebellum period represented a minority within 
the white population, and ownership had clearly become concentrated relative 
to its diffusion within the white population in the colonial period. On the 
other hand, the racial balance between whites and blacks tended to equalize 
during the first part of the nineteenth century. 

After the Civil War 

St. Stephen's in the colonial and antebellum periods was an element in the 
low country economy that saw Charleston emerge as its economic hub. Charles- 
ton's prosperity rested on a commercial system consisting of three elements: 
commodities produced locally or brought to the city for export; transporta- 
tion facilities of roads, canals and railroads linking Charleston with the 
interior; and a first-class natural harbor (Moore 1979:156). The effects of 
westward migration were not felt immediately in the city, although by the 
1850s, Charleston's population saw an absolute decline despite growth in the 
urban population nation-wide and the population of other southern cities 
(Steen 1970:38-39). 

This configuration changed entirely after the Civil War. No manufacturing 
base emerged in the vicinity of the city, and the development of the textile 
industry in the upcountry during the closing decades of the nineteenth and 
early decades of the twentieth centuries enhanced the antebellum difference 
between the coast and the Piedmont. Industrialization of the upcountry pro- 
vided an export base for a regional economy defined today by the Greenville 
and Spartanburg Standard Metropolitan Statistical Areas (Moore 1979:165). 
Railroad development in the post-war period bypassed Charleston, with the 
majority of South Carolina cotton shipped via Norfolk (Moore 1979:161). 
Sustained economic revitalizatlon of the city began at the turn of the cen- 
tury as a result of improvements in naval installations and additional feder- 
al projects (Moore 1979:170). Cotton prices were low through the end of the 
nineteenth century, and although they had begun to turn back upward in the 
early twentieth century, the dislocations caused by World War I, depredations 
of the boll weevil, and the Depression finally destroyed the cotton market 



40 



(Woodward 1951:181-182; Tindall 1967:121-122, 428-429; Table 6). Crop di- 
versification projects during the New Deal ended the farmers' dependency on 
cotton (Tindall 1967:403-406, 428-430), and presently the farms in the vici- 
nity of the town of St. Stephen produce soybeans, corn, sorghum and some 
cotton. 



41 



TABLE 6. Berkeley County, Cotton and Corn Production 

1889-1939 



Year Cotton (bales) Corn (bushels) 

1889 12,557 296,528 

1899 10,419 368,400 

1909 17,415 391,195 

1919 10,867 468,772 

1929 2,944 367,197 

1939 3,176 347,885 



Sources: U.S. Census: 1895 , pp. 383, 396, 407. 

U.S. Census: 1902 , VI, Pt. 2, pp. 181, 433, 436, 489. 

U.S. Census: 1913 , VIII, p. 516. 

U.S. Census: 1922 , VI, Pt. 2, p. 287. 

U.S. Census: 1932 , II, p. 481. 

U.S. Census: 1942, I, pp. 461, 477. 



42 



IV. TESTING PHASE 

Introduction 

The archaeological testing phase began in St. Stephen on March 12, 1979 and 
was to include test excavations, surface collections, metal detector survey, 
mapping, and limited historical research on five sites. These sites were 
38BK73, 38BK75, 388K76, 38BK88, and 38BK91. Site 38BK245 was not tested by 
SSI, but was included in the mitigation phase. From the survey data, pre- 
liminary interpretations of site function were made and used to guide field- 
work until testing data were sufficient to make educated field decisions. 
Concurrent with the field testing, a field laboratory was set up in quarters 
provided by the Corps of Engineers at the project headquarters. All material 
retrieved from the field operations was to be washed and preliminarily cata- 
logued in the field laboratory, in order to provide immediate feedback for 
informed field decisions. The testing phase was successfully completed on 
April 6, 1979, four weeks after it began. As a result of fieldwork, document- 
ed recommendations were developed and submitted to IAS-Atlanta on April 18, 
1979. A contract for mitigation at three of the tested sites plus one addi- 
tional site was negotiated and signed, and mitigation of the four sites began 
on June 14, 1979, seven weeks after termination of testing. Mitigation 
fieldwork was completed on October 19, 1979 and is discussed in Chapter VI. 

The following sections describe the hypotheses, methods, results, and recom- 
mendations of the testing phase and, more importantly, describes the data 
which caused a major shift in hypotheses developed for the subsequent mitiga- 
tion phase. It should be noted that the mitigation effort for site 38BK73 is 
included here since the site ultimately proved to be a treefall and field 
scatter, and offered no data pertinent to the mitigation goals of the 
project. 

Testing Goals 

The Request for Proposals developed by IAS-Atlanta raised many questions of a 
general nature concerning the sites and historical archaeology. The ques- 
tions can be grouped into three categories: Those dealing with the ethnic 
identity of the inhabitants and their social stratification; the settlement 
and economic relationships within the region; and, lastly, the context of the 
sites within the larger framework of the field of historical archaeology 
(Garrow 1979a). Only some of these questions are listed here to illustrate 
the range of research required and to minimize redundancy. 

Under the subject of ethnicity and social stratification, IAS-Atlanta listed 
the following questions: 

1. What 1s the social identity of the Inhabitants of the sites? 

2. Is it possible to identify Huguenot as opposed to British American 
behavioral patterns? 

3. How much and what kind of interaction and social stratification was there 
between French and British Americans? 



43 



Under settlement and economic relationships were listed: 

1. Is it possible to determine plantation economics, including domestic vs. 
imported goods, the role of Colonoware, and the economic base of the plan- 
tations through archaeology? 

2. What is the relationship between the plantations and other plantations, 
the town of St. Stephen, inns, churches, craftsmen, and factors? 

3. How do the economics, settlement, and social relationships of these sites 
compare to other sites 1n the area? 

4. How do such relationships change through time in the Santee River area? 

Under the general category of the sites' contribution to historical 
archaeology were: 

1. What is the overall significance of these sites to contemporary archaeolo- 
gical knowledge? 

2. Are these sites representative of Huguenot plantations? 

3. Can portions of a system be considered as representative of the whole 
plantation system? 

Since the data provided in the scope of work did not reveal whether any 
structural remains existed at the sites or whether trash features had been 
found, it was proposed by SSI to place the major thrust of the testing phase 
on locating structural remains, features, and artifact patterning to 
determine whether 1t was even possible to begin testing more sophisticated 
questions. Along with this archaeological testing program, preliminary his- 
torical research was proposed to determine geneaological data and land owner- 
ship, and to assess the available records for their capacity to answer the 
questions posed by IAS-Atlanta. 

Specifically, it was proposed for the testing phase "to conduct an orderly 
data collecting process that can result in pattern recognition during analy- 
sis"; "to locate architectural ruins" and trash features; and finally to 
analyze and assess the significance and overall potential for further data 
recovery at the sites (Garrow 1979a). Archival research was to be directed 
towards developing a concise economic history of French Huguenot settlement 
in the area; construction of chains of title; compilation of comparative 
French-British economic data in the area during the period under study; and 
lastly, locate plantation papers, diaries, and wills generated by the inhabi- 
tants of the sites. 

While these goals are not hypothetlco-deductive in nature, more research was 
needed to determine whether or not more specific questions could be asked, 
much less answered. It should be noted that the historical research 
conducted during testing concentrated on Yaughan plantation, the location of 
Sites 38BK73, 38BK75, 38BK76. This research dealt mainly with the chain of 
title and geneaology of the Cordes family, who owned the plantation during 



44 



the eighteenth and early nineteenth centuries. However, the time required to 
locate and interpret the data took longer than anticipated, because the 
amount and accuracy of the research already completed prior to testing had 
been overestimated. As a result, no substantive efforts could be made 
towards establishing chains of title at the other two sites (38BK88 and 
38BK91). As testing progressed, it became obvious that these two sites would 
not contain enough archaeological integrity to warrant mitigation. 
Discussions of the historical research from both testing and mitigation are 
presented in Chapters III, V and XI of this report. 

Sites Recommended for Clearance 

Site 388K91 (Figure 5) was located in a "pine plowed" field which had been al - 
lowed to revert to pine within the four years prior to testing. Such clear- 
ing of a pine forest involves not only cutting down the trees and raking back 
the brush, but also removal of the stumps and larger roots. It is not sur- 
prising, therefore, that this site failed to produce any in-situ features. 

Due to the size, nature of the site, and ground cover, this site was not 
gridded. All test pits were laid out north-south along a line through the 
center of the site, as determined by surface collections of artifacts, which 
in this case were predominantly brick fragments. The surface collection was 
augmented by shovel cuts and a brief metal detector survey. 

The typical stratigraphy at Site 388K91 (Figure 6) consisted of approximately 
six inches of brown sandy topsoil overlying four to five inches of light yel- 
low sand, resting on red clay subsoil. Except for Sites 38BK88 and 388K76, 
this stratigraphy was typical for the other sites tested (i.e. brown sand, 
light yellow sand, red clay subsoil). Site 38BK88 had yellow clay subsoil in 
place of the more typical red clay, and Site 38BK76 had an additional layer 
of humus overlying the brown sand. 

The artifacts recovered by excavation were principally brick and charcoal 
fragments and, except for surface finds, the artifacts were found in the 
brown sand layer. Only 11 datable historic sherds were found from surface 
and excavation combined. The mean ceramic date is 1836.4, but this date is 
undoubtedly affected by sample size. Three test units were dry screened with 
three-eighths inch mesh (Test Units 1, 2, and 5) and three were hand sifted 
(Test Units 3, 4, and 6). It is not felt that hand sifting greatly affected 
the relative frequencies or totals of cultural material collected at this 
site. 

Vitrified brick or clay fragments and a few lumps of metal, probably nails, 
were found along with the great amounts of brick and charcoal fragments. 
Because of the low amount of ceramics and high percentage of brick fragments, 
it is hypothesized that Site 38BK91 represents a brick kiln or clamp. 
However, the effects of clearing operations and the lack of features during 
testing indicated that there was a very low probability of recovering any 
additional data to substantiate this hypothesis. Therefore, no further work 
was recommended at Site 38BK91. 



45 



\ 



\ 



\ 



\ 



\ 



\ 



\ 



Test Units 



Oaturn 



© Corps of Engineers Benchmark 




T. U. 6 



\ 



Site Boundary 



CORPS 
Is B - 5 - 2H97S c 



/ 



njLJ 

) 15 30 

SCALE IN FEET 



FIGURE 5 
Site 38BK91 
Testing Site Plan 



46 



Ui- 
00 



LU 

.J 
U. 

o 
or 

0. 



00 

< 

LU 



iLi- 









SSsKs 








(A) 



oo" 



LU 

_J 

U- 

O 

or 

0_ 









11^ K* 



SIP 














</> 






0) 






/— 






• ^ 






<4- 






o 






- 






C_ 




1— ■« 


_ 




CT1 






^_ 


-t-> 


U3 


cc 


■^ 




cc 


c 


LU 


m 


=1 


Q£ 






— 


OJ 


— 1 


CJ 


+-> 


(/> 


^-* 


•f« 


O) 


u_ 


GO 


1— 



LU 
LU 

U. 






T3 








C 




a 




en 


T3 


5 


C 


o 


o 


= >% 


CO 


jo a 


c 


>- — 


3 

o 

— 




03 


_j cr 


5§p 


F^"l ( 




^Hs 


P •■•*■ 






££& 


i£ 







47 



Site 38BK88 (Figure 7), with a mean ceramic date of 1808.8, was probably an 
isolated small farmer's house, possibly with a few small outbuildings. No 
conclusive statements concerning intrasite function could be made at this 
site since it had been deep plowed to ±24 inches below grade for ten years 
prior to testing (Gaillard 1979, personal communication), destroying all but 
the deepest features and disturbing subsoil to a depth of one foot. 

The site was located next to the present Lake Moultrie, but according to the 
1921 USGS Chicora Quadrangle map, the nearest large topographic feature to 
the site before the lake was constructed was Buckhall Bay to the southwest, 
which appears on the 1825 Mills map of the area as Buckhall Swamp, and on the 
1773 Mouzon Map as an unnamed bay or swamp. No houses are recorded on any of 
the maps at the location of Site 38BK88, although a 1943 USGS Quadrangle map 
does show two houses near the site on the edge of Lake Moultrie, neither of 
which is the site in question. 

Site 388K88 can be characterized as being located in swampy uplands and was 
probably surrounded by dense pine barren at the time the structure was con- 
structed (Herold and Knick 1978). It was approximately three miles inland 
from the edge of the Santee River floodplain, and from location alone it 
seems certain that the site was not one of the most desirable plots staked 
out by the original European settlers of the area. Although it is possible 
that the site's inhabitants cultivated inland rice fields (Herold and Knick 
1978), it seems more likely that they were subsistence farmers in a rela- 
tively marginal agricultural zone. 

Since the site was nearly level and totally free of ground cover, it was 
gridded into 50 foot squares from a north-south baseline. Each 50 foot 
square was Intensively (100 percent) surface collected and all grid lines 
were tested with a metal detector. 

Controlled metal detector surveys at all sites involved checking for metal 
within a four or five foot wide swath along each grid line. Upon location of 
metal, a wire surveying flag was placed in the ground at that location. Once 
all of the metal was flagged, the Hags were mapped on the site map. Concen- 
trations of metal on Figure 7, noted by solid lines, were determined on the 
basis of one or more flags per linear foot of grid line. The broken lines on 
the figure indicate two or more flags per five linear feet of grid line. 
There were also isolated indications of metal elsewhere on the grid but these 
were often 25 to 50 feet from each other and did not form concentrations. 
The information developed from the surface collections and metal detector sur- 
vey was plotted on maps, and six 3 x 3 foot test units were placed where it 
was determined they would produce the most subsurface information. 

As noted on Figure 7, six 50 foot grid squares contained over 20 "datable" 
ceramic types (South 1977a: 210-212 and Noel Hume 1978) and nine contained 
between 5 and 20. Most of the remaining squares had none. The six most 
heavily concentrated squares were also in areas of high metal density. Based 
upon this Information, this area appeared to have the highest probability of 
yielding features and, for this reason, test units were placed there. The 
first test unit exposed the remains of a well, probably the only feature left 
at the site, and the only one found during testing or mitigation. 



48 



3- 



3- 




5« 


5§ 


i ■ 

*2 


4 2 




iS 


jSj 


->■* 


*a 




$ 


£u 


Sfc 


** 




A" 


} 3 


5 
















£ j 


I* 


UIO 


■±?o 




>f 


»•* 


»-<r 


►* (X 


»_ 


S" 


"I 


$ z 


St z 




3* 

MS 


I 1 ^ 


if 


5^ 

« z 
z o 


£ 



tk 


un 


o z 


r^ 


in — 


^ C 


1 


P«» CO <T3 


< 


CO r- 




uitno. 




c; 




3 IV 91 




CD +J 4J 




l— i -r- -f— 




U_ (/■> t/) 



+ 



3- 



2h 



■f- 



fejSj,;;;.; 


::••::•*" 


::.';;;;;,:;, 


:.:■;;; 






|i;:;::>;:*< 


: ^- 








49 



It is possible that squares 50N/300W, 50N/200W, and 150S/200W (northeast 
corner coordinates) represented one or more outbuildings or trash features, 
but a definite conclusion could not be made because of the lack of data and 
subsurface integrity. 

The typical profile at 38BK88 was approximately six inches of brown sandy 
soil blending with an underlying layer of light yellow sand. The six inch 
thick layer of light yellow sand then blends with light yellow clay at the 
bottom, where deep plowing had greatly disturbed the subsoil. Ninety-eight 
percent of the artifacts recovered from the test units, excluding brick and 
charcoal, were obtained from the brown sand layer, even though all levels in 
all pits were thoroughly dry screened with three-eighths inch mesh. 

Sites 38BK73, 38BK76, and 38BK91 showed that no features were recorded deeper 
than two feet below the surface at the sites during testing. This depth of 
features was further supported during mitigation, since only a few features 
at 38BK75 and 38BK245 extended more than two feet below the surface. The 
deepest trash or clay extraction pits at Site 38BK75 occasionally extended to 
three feet, and at 38BK245, one cellar extended approximately three feet and 
another four feet below the surface, out of a total of several hundred 
features. 

It seemed safe to assume that if any features were present at 38BK88 they 
should have begun in the second level, light yellow sand, and continued into 
the top of the subsoil. Since plowing had disturbed the soil to an average 
depth of 24 inches, this means that on the average the subsoil was disturbed 
to a depth of at least one foot below its interface with the light yellow 
sand layer. On the basis of this information and a lack of any features 
except the well, it was clear that only the deepest features, wells and 
privies, would be left at the site. 

The only feature that had probably survived at the site was the well found in 
Test Unit 1 (Figure 8). After clearing the disturbed overburden from the 
well, rotted brick stains began to show at the base of the unit. As may be 
seen in the figure, the outline of the well was fairly clear below the 
overburden-subsoil interface. The well was augered to 4.15 feet below the 
ground surface (or 2.1 feet below the bottom of the unit) using a three inch 
hand- turned bucket auger. Augering was discontinued when the boring began 
caving in, and the very wet clayey sand filling the well continually slipped 
out of the auger before it could be retracted. The auger did not produce any 
artifacts, but it did show conclusively that the feature was a well filled 
with sand and some clay and that it still acted as a retaining vessel much 
1 i ke a dry wel 1 . 

Before fieldwork began, IAS-Atl anta had hypothesized that 38BK88 was the re- 
mains of a tavern noted on the Mouzon map of 1773. Since taverns served 
wine, among other things, and wine is shipped and served in barrels and 
bottles, one should have expected to find a relatively higher percentage of 
olive wine bottle glass at a tavern than at a purely domestic site. The 
amount of olive glass, as a percentage of Kitchen Group artifacts at 38BK88 
from all contexts, was 5.7 percent. At South' s ( 1977a: 126-127 ) tavern site 
(Brunswick 25), the percentage was 17.3 percent, and was generally over 10 
percent at all of his sites. At the slave sites on Yaughan Plantation, the 



50 




LO S_ 

*C"OW 

C C <T3 

(T3 (T3 <U 

l_ 

.— I C 

00 fO M- 

CO CO i— O 

i<£ +j a. 

00 CD ■<- to 

CO C O) <D U_ 

uu ro =) c i— «r 
=> tU -t-> 3 M- i— i 

e: -m to o o i 

«— ■ -t— <U -C i. CO 

U- OO I— CO C 00 



51 



percentages were also low: 8.9 percent at 388K76 and 9.6 percent 388K75. 
Curriboo Plantation, the percentage in the slave quarters was 14.2 percent. 



At 



Further, it is interesting to note that at Yaughan and Curriboo, the percent- 
age of wine bottle glass obtained from surface collections was always con- 
siderably higher than that from excavation. If this dichotomy holds true at 
Site 38BK88, the overall percentage of wine bottle glass should actually have 
been lower than 5.7 percent, since most artifacts at that site came from the 
surface. From data gathered on olive green wine bottle glass alone, it would 
appear that Site 388K88 was not a tavern. 

The nearest tavern on the Mouzon Map (1771) to 38BK88 was approximately one 
mile to the northeast of the site. While historic maps often have a tendency 
to be unreliable in some respects, this does not appear to be the case here. 
This same map correctly locates the St. Stephen church, the Santee River, all 
major roads, creeks, and swamps to a tolerance of several hundred feet when 
compared to modern maps of the area. A mistake of over a mile in the loca- 
tion of a tavern seems unlikely in this case. 

The mean ceramic date of the site is 1808.8 and the occupation range is 
1790-1827 (South 1977a: 215). This puts the site 17 years too late 
(1790-1773=17) for the 1771 Mouzon Map. Although this method of dating oc- 
cupation ranges is inconclusive, only one ceramic type has a beginning date 
even close to the 1771 Mouzon Map and that is lighter yellow creamware, which 
Noel Hume and South (South 1977a: 212) begin dating at 1775, four years after 
the Mouzon Map. 

In summary, Site 388K88 was too late, too far away, and had too few wine bot- 
tles to be the tavern noted on the Mouzon Map. Our interpretation of the 
site, as noted above, is that it was probably an isolated farmstead. We did 
not recommend mitigation at Site 38BK88. The primary reason for this recom- 
mendation was that the proposed impact from dumping activities could not 
impact the site any more than deep plowing had already done. Furthermore, 
even if dumping activities were to have included complete excavation of the 
site, the expected return from any mitigation excavations would not have 
produced much more information about artifact distribution and patterning 
than was already available from the testing program. 

Sites Recommended for Mitigation 

Site 38BK73 (Figure 9) was located in an open field with moderate ground 
cover approximately 275 feet to the west and four to six feet in elevation 
above a small creek. To the west of the site was a dirt road leading to 
Sites 38BK75 and 38BK76 to the northeast. This road may have been the main 
access road to the principal buildings of Yaughan Plantation, although there 
is no documentary proof for this. Until moved to the west by the Corps of 
Engineers, the road was used by the Piatt family to get to their homes beyond 
Sites 38BK75 and 38BK76. 



52 



en 

m c 

O"! CO (O 

CO r— 

LU CO Q_ 

C +J w 
>— < .^ .,_ 
lu u-> ts) 




0v Qh 



•+ 




I 
I 



H- 



? 



UJ 

UJ 



« < 

N o 



4 



t 



I 
9-f- 



H- 



:+■ 




20 ? O 

5 ° 



"2 2 2 



53 



Site 38BK73, located by Brockington (1980:35) during his survey, was des- 
cribed as a thin scatter of late eighteenth to late nineteenth century arti- 
facts. Because it did not produce any subsurface artifacts from three shovel 
tests, Brockington did not recommend any testing at the site, although he did 
recommend that the site be monitored during construction. His reasoning was 
that even though the site did not produce subsurface artifacts or features, 
it was one of only four upland eighteenth century sites found during the sur- 
vey and, as such, was rare in the project area. 

It was later decided by others (including the President's Advisory Council) 
that the site should be tested. Seven 3X3 foot test units were excavated 
and the soil dry screened with three-eighths inch mesh. A 50 x 50 foot grid 
was laid out, each square was surface collected, and the grid lines were 
submitted to a metal detector survey (Figure 9). 

The results of the surface collection and metal detector survey were disap- 
pointing. Only six artifacts were recovered, compared to over 30 found dur- 
ing the survey phase (Brockington 1980), and the metal concentrations were 
not as extensive or as dense as those at 38BK75 and 388K88. 

The test units produced more artifacts than the surface collection, 36, but 
these did not help very much in site interpretation. Test Unit 5 eventually 
produced the remains of a picnic or trash fire of very recent origin, and the 
metal noted in the vicinity of Test Unit 5 on Figure 9 is undoubtedly the 
remains of this same feature. 

As testing was being completed at 388K73, a concentration of what appeared to 
be brick fragments was noticed outside the gridded area. In order to deter- 
mine whether these were the remains of a structure or chimney fall, two test 
units, numbers 6 and 7, were excavated. Upon completion of the test units to 
a yellow clay feature, auger tests were extended to the south and east from 
the sides of Test Unit 5 to establish the limits of the feature. The result 
of this operation was the diagram in Figure 10, showing the estimated extent 
of this feature. Based on this data, it was recommended that 38BK73 be 
accorded the lowest priority for the mitigation, after 38BK75 and 38BK76. 
Specifically, it was recommended that a 20 x 20 foot block be excavated 
around the feature (Garrow 1979a). 

Mitigation began in June 1979 and lasted one week plus two days with half a 
crew. During this time, a 20 foot block was excavated in five foot square 
units with one foot baulks left between units (Figure 11). All soil was dry 
screened through 1/4-inch mesh and soil samples were taken of representative 
natural levels and units, as well as extensive samples of the yellow clay 
feature itself. There was no evidence of deep plowing, although plow scars 
were visible 1n the top inches of the yellow clay feature and surrounding 
soil (Figure 12). 



54 



NO 
CLAY 

CLAY 


K 




5o|o 


Z 

so 


z 

< — 


jJQj 


2*- 


Jj*Z 


idyui 


<< 

(El 


d £ 


>>>> 


U.H 


K 




Z 


*z 


a: <r 


V> UJ 


nw 


UJ UJ 


y 


O o 


-z 


|2 


= 3 


tro 


UiO 


< < 


(StJ 


>u 




■O 
c 
n: 

o a* +J 

— * cc ■— 
cz c 
L4 n 2 
a: 
z: ai — 

C^ -M 1/1 

•— i -«— aj 

L_ U-, (— 



55 




MITIGATION AT 
SITE 38BK73 
LOOKING NORTH 




I ^1- ^" "'" -* --*"*■ -•• III 



TESTING AT 
SITE 38 8K75 
LOOKING NORTH 



FIGURE 11 
Photos of 
Site 38BK73 and 38BK75 



56 



now 



I03W 



IOOW 

— I — 



93 W 

— 1 



90W 




WATERLAID 3ANO 

oemessioN 



30 N 



43 N 



40 H 



- 33 N 



30 H 



N 




ruin 

3 

SCALE IN FEET 



$88?! 



f>'i"'Vi" 



Primary Burnt Area 
Secondary Burnt Area 
Subsoil Pulled up by Tree 
Roots 



FIGURE 12 

Site 38BK73 

Block Excavation Tree Fall 



57 



Once the feature had been nearly cleared off it became evident that no post- 
holes or other architectural features were present. It was then hypothesized 
that the feature, which showed evidence of having been burned, represented a 
wattle and daub chimney fall covering all other architectural evidence. The 
lack of artifacts, less than had been found during survey or testing (only 
14), further indicated a special use structure such as a tobacco barn. As 
the last 5 foot unit was excavated, what appeared to be a post and charcoal 
concentration seemed to support the hypothesis of a non-domestic structure. 

However, once this unit was completed, the baulks removed, and the field map 
critically examined, it became obvious that 38BK73 was a treefall that had 
burned in place. The "post" and charcoal were the incompletely burned roots, 
the fired clay was the area under and around the tree trunk and larger branch- 
es, and the yellow clay "floor" or "chimney fall" was deeply buried subsoil 
which had been pulled up by the tree roots, and not imported from elsewhere 
as had been suspected. 

The presence of artifacts, which were most heavily concentrated on or near 
the surface of the field, was most likely due to trash dumping activities by 
the inhabitants of Site 38BK110, an eighteenth-nineteenth century site out- 
side the boundaries of the project, and located to the north of 38BK73. It 
is possible that this corner of the field was used by others for dumping as 
well . 

Site 38BK75 (Figures 11 and 13) was found by Brockington (1980:45-47) during 
TTTs survey and described as a heavy concentration of eighteenth and early 
nineteenth century artifacts in two loci. Locus B was outside the project 
area, and a surface collection there during testing indicated that it was 
probably the site of the main plantation house. Locus A was a heavy con- 
centration of eighteenth and early nineteenth century artifacts located in an 
open field next to the Piatt access road and included a heavy concentration 
of Colonoware. Brockington recommended testing and mitigation at Locus A. 
Testing was conducted and bore out Brockington' s high expectations for the 
site. 

The testing program extended over a period of one and half weeks with a two 
person crew. A 50 foot grid was laid out along a slight ridge which runs 
northeast-southwest through the main portion of the site. Testing subse- 
quently showed that Lod A and B were not connected while Sites 38BK75 and 
388K76 to the east were connected at the northeastern end of 38BK75 Locus A 
and the northwestern or western end of 38BK76, by an area almost twice as 
large as the original 388K75 alone. No attempt was made to place test pits 
in this intervening area, as the scope of work indicated the two sites were 
discrete, and the level of effort was for two medium sized sites and not 
three and a half. For these reasons, the concentration of artifacts at 
38BK75-Locus A (Figure 13) was gridded and surface collected. A metal 
detector survey was run following all grid lines. After plotting this 
information on a map and field inspection of the site's varying elevations, 
16 test units were placed where they promised to produce the most informa- 
tion. Two features were located, a small disturbed Archaic Period lithic 
scatter was recovered in Test Unit 5 in the light yellow sand layer, and a 
historic trash pit in Test Units 1 and 1A. Since the major thrust of the 



58 



t|- 



-t- 



-i- 



-^ 






LD 






r~. 




en 


--*!. 


c 


I— 1 


CD 


(0 




00 


r— 


UJ 


co 


Q. 


cs 






ID 


a) 


0J 


O 


+-> 


+J 


i— • 


•^ 


•f— 


u. 


00 


U~) 




5* 

" O 

I 5 



IS 



Ha M " ? 



:|= 



51! 



59 



testing program was the Historic Period, and the prehistoric artifacts re- 
covered were sparse and disturbed, it was decided that the prehistoric com- 
ponent did not warrant further investigation at that time, and work was 
directed towards investigating the historic component, especially the trash 
pit. 

Test Unit 1 was taken down to a point 1.3 feet below the surface where the 
feature appeared to be square (Figures 14 and 15). Test Unit 1A was taken 
down approximately six inches below the surface to the top of the same fea- 
ture to help determine its horizontal extent. In conjunction with excavation 
and in the interests of preserving the feature, auger borings were placed 
across the feature to determine the horizontal and vertical dimensions 
outside the test unit without excavating and thereby damaging the feature. 
This proved partially successful and resulted in a hypothetical outline of 
the feature (Figure 13), its depth and, combined with screening of feature 
fill in Test Unit 1, its contents. 

The typical profile at Site 38BK75 was similar to those at 388K73 and 388K91 
(Figure 15). Eighty-five percent of all artifacts, excluding brick and char- 
coal , were found in the brown sand layer. The presence of the preserved fea- 
ture at the site shows that ground disturbance of the historic component had 
been superficial, and the probability of in- situ remains was high. The site 
appeared to consist of a series of artifact concentrations running northeast 
to southwest along the center squares of the grid (Figure 13). A summary 
discussion of the testing results at 38BK75 and 388K76, the Yaughan slave 
quarters is given after the following discussion of Site 38BK76, 

Site 38BK76 (Figure 16) was approached differently from the other four sites. 
This was due to several factors: large piles of brush produced by partial 
logging of the site, the large size of the site, the many remaining trees and 
heavy root systems, and restricted access to and within the site produced by 
high groundwater (Figure 17). 

It was physically impossible to set up a grid within the time limits imposed 
on the project and a decision not to use damaging heavy machinery on the site 
during testing. It was, therefore, decided to use a series of raked back 
areas, where vegetation permitted, to expose as much of the surface as pos- 
sible. Thirty-five rake backs were placed, primarily on the southeast, 
south, and southwest quadrants of the site. These were then visually in- 
spected for artifacts and mapped (Figure 16). Rake backs 1, 14, 18, and 34 
produced a total of seven artifacts of which three were in the largest rake 
back (number 1); two were prehistoric, and five were historic. 

Following the rake backs, the three roads crossing the site on its northern 
half were inspected to determine the northern and east-west extent of the 
site. It was during this phase that the connection between Sites 38BK75 and 
38BK76, as noted above, was discovered. This connecting area was heavily 
rutted by the roads and all low spots contained standing water. 



60 



11 



33 



553 



JE3 ;gKft Cj a B 



l I 



^^ssssmsom 



9 -65 



_^_ zn m> m &%*5Z 



z & Ts5 cG3b±x!±rki 



2 c 



Ennm 





I— C 


in 

«3" r*~ 


1—1 






i— i ^. 


4-> 






en 


•r— 




UJ CO 


C 




13 .. 


oc en 


^_ 




< * 


=5 






< 


C3 0) 


+■> 




U. 


►—i +-> 


(/I 


o 


o « 


U- •!- 


CD 


1 


u. c r 
o 7a 


GO 


1— 


L.o 


1 :^ 








SSs 






■« 


>- ie 






c 


< .1 

Q. E _ 






■ o 


5?: 







as 



Ei-zaoa 



I r 

31 

« 
5 



~" I ' i 



W * & ? E 




ssp^r^g y ^g q 




QiilBIlIi 



TEST UNIT 



TEST UNIT IA 



S W 



61 
NW 






Unexcavated 



WEST PROFILE 



SE 



SW NW 







NORTH PROFILE 



SOUTH PROFILE 



TEST UNIT 




EAST PROFILE 



RA11 T 



A' 






' ■-■••-•' 



0.3 I 

SCALE IN FEET 



Brown Sand 

Brown Sand, Brick, and Charcoal 
Light Yellow Sand 
Brick Stain 
Red Clay 



°"^'^"''jv ■-^'■"•'"•'■''i.r-Y'v v-i-y ■ ■■'•'■ ' '. - ' ;■ ■"" ""•'"' 'iV' * •■'--'■"■■ ■•:.. 
v.iTr.\-:'.ri'X-'-' ■[•NO:'" •.-" : '_>*■.- " <>•■ '-.■•■■ '. ;;" : ' ':>--v" # "-•■•••■!-? - 



i 1 .". 1 1-!,' . ■ .&;.-, .;■■"?■■ -. .VW- ■-.-•■• niw ."V. ^.'" 



CENTER PROFILE 

FIGURE 15 
Site 38BK75 
Profiles-Test Units 1-1A 



62 








- 






"3 






^» 






C 






<D 




lO 


— i 




r^ 


■flH 


vo 


^ 


t/1 


f— 1 


03 






CO 


en 


LJ 


m 


c 


CS 






3 


O) 


w 


CS 


•M 


(/) 


>— i 


•^» 


Ol 


u. 


t/1 


1— 



o ' 



£ UJ 



o < 



. — § 



63 




TESTING C0N0IT10NS 
AT SITE 38 8K 76 




CONSTRUCTION ANO 
SORROW PIT AT SITE 38 8K245 
LOOKING WEST PRIOR TO 
MITIGATION PHASE 

A - STRIPPED AREAS 

8 - MOUNO LEFT AROUNO KILN (243K) 

C -AREA OF BRICX STRUCTURE (245C) 



FIGURE 17 

Photos of Site 388K76 

and Site 38BK245 Testing 



64 



Once the limits of the site were more or less determined, 15 test units were 
placed to more precisely locate the site boundaries (Test Units 7, 9, 11, and 
14), to understand the stratigraphy and check its depth (Test Units 1, 2, and 
3), to search for probable areas of buried structures following a metal detec- 
tor survey (Test Units 5, 6, and 8), and to locate features in areas consi- 
dered to be high probability areas because of elevation and topography (Test 
Units 4, 10, 12, 13, and 15). Of these test units, numbers 3, 8, 13, and 14 
produced features. There were two postholes in 3; a single posthole in 8; a 
possible posthole in 13; and a possible posthole in 14. For future reference 
the postholes in Test Unit 3 were part of Structure 76A and the posthole in 
Test Unit 8 was part Structure 76B. The possible posthole in Test Unit 13 
was not associated with a structure, and the possible posthole in Test Unit 
14 may have been associated with Structure L. 

The typical stratigraphy at Site 38BK76 is illustrated by Test Unit 3 (Figure 
18). Approximately 1 1/2 inches of humus and root mat overlies six inches of 
brown sand over six inches of light yellow sand, resting on red clay subsoil. 
Over two- thirds of the historic artifacts, excluding brick and charcoal 
fragments, were found in the humus and brown sand layers. The prehistoric 
(mainly Woodland) artifacts were found in these top two levels as well. No 
prehistoric features were located. The stratigraphy of the entire site 
appears to be relatively undisturbed since occupation in the early nineteenth 
century. Local informants maintain that the site was not logged before the 
present partial logging. Although this is undoubtedly in error, it does seem 
certain, from the predominance of hardwoods, that the site has not been great- 
ly disturbed for the last 75 to 100 years. For this reason and the presence 
of five intact features located during testing, the probability of delineat- 
ing structures and associated features was high. 

Because of the sparseness of the prehistoric component, its restriction to a 
small part of the site and its position across a creek from and to the east 
of 38BK236, of which it probably formed a minor part, no emphasis was placed 
on this component during the following mitigation phase. 

Analysis of Sites 38BK75 and 38BK76 

Since it was clear at the end of testing that Sites 38BK75 and 38BK76 were 
connected, it was felt that they should be discussed and analyzed together. 
A large collection of ceramics was retrieved from Site 38BK75, Loci A and B. 
Testing at Site 38BK75-Locus B were restricted to controlled and uncontrolled 
surface collections, as that locus was outside the project boundaries. The 
ceramics retrieved from 38BK75-Locus B are presented in Table 7. 

The site description in the Request for Proposal and Brockington's (1980) 
report stated that 38BK75-Locus B represented an overseer's or master's 
house. The ceramic frequencies cited above in Table 7 reinforce that 
interpretation, but conclusive proof in the form of historical research has 
not been found. It is probably significant that colonoware is represented as 
a mere 2.3 percent of the total ceramic inventory at Locus B. If the site 
was an overseer's or master's house, the occupants probably would not have 
been using or generating colonoware. 



65 




WD CO 
I — 
00 S<£ +J 

<— i CO •«— 

co c 
IxJ <**) S 
a: 

CD +-> </) 

•— i -t— O) 
U. </1 t— 







fi£$8 

* * ■» %£$$&■? --Ill 



66 



TABLE 7. Historic Ceramics Recovered from 38BK75-Locus B (Testing) 

Site 38BK75-Locus B Total Surface 

Non-local Ceramics 

Creamware 53 10.9% 

Pearl ware 285 58.5% 

Porcelain 19 3.9% 

Whiteware 99 20.3% 

Delft 0.0% 

Stoneware 18 3.7% 

Redware 0.0% 

Soft Paste Earthenware __2_ 0.4% 

476 97.7% 

Local Ceramics 

Colono 11 2.3% 



67 



The ceramic frequencies from Locus B contrast sharply with those recovered 
from Locus A, as presented in Table 8. Locus A was interpreted as a 
habitation site occupied by African slaves on the basis of the large 
percentage of Colono sherds recovered. There is little doubt that Loci A and 
B overlapped in time, although Locus A carried a mean ceramic date (South 
1977:207-218) of 1803.0, and the mean ceramic date of Locus B was 1814.3, as 
determined by testing phase material alone. Ultimately, Locus A had an MCD 
of 1789.8. Time lag (Adams and Gaw 1977) of 24.5 years (1789.8 to 1814.3) is 
hardly surprising between an overseer's/owner's house and slave habitations. 

As described above, Site 38BK76 presented severe collection and testing dif- 
ficulties. Despite these difficulties, 38BK76 yielded the largest ceramic 
collection of the three sites recommended for mitigation. Table 9 presents 
the ceramic sample extracted from this site. 

The most striking feature of the 38BK76 testing collection was the over- 
whelming amount of colonoware. At 38BK75-Locus A colonoware sherds accounted 
for 51.2 percent of the sample of sherds. Site 38BK76 yielded a slightly 
larger sample of 556 sherds, of which 522, or 93.9 percent, were colonoware 
sherds. Assuming that colonoware sherds were made and used by slaves, it 
followed that the residents of 38BK76 had less access to manufactured goods 
than did the residents of 38BK75-Locus A. This factor seemed, on the basis 
of testing, to indicate that the residents of 388K76 occupied a lower 
socioeconomic status within the plantation system than did the slaves that 
lived at 38BK75-Locus A. It was suggested that the status differential 
observed between these two areas represented the dichotomy of house and field 
servants within the plantation system. 

Study of the relative topography of Sites 388K76 and 38BK75, Loci A and B, in- 
dicated that 38BK76 occupied the area closest to the Santee River swamps and 
as such the lowest elevation of the three sites. Figure 19 depicts the aver- 
age elevation of each site and the percentage of local and non-local ceramics 
recovered during testing. This figure clearly indicates that 38BK75-Locus B 
not only occupied the highest elevation among the three sites, but also 
contained the highest percentage of non-local ceramics. It was felt at the 
end of testing that relative elevation and distance from the Santee River 
might be indicators of status differences. 

The mean ceramic date (South 1977a: 207-218) achieved from 38BK76 after test- 
ing was 1786.8. That date is 16.2 years earlier than the testing date 
achieved for 38BK75-Locus A and 27.5 years earlier than the testing date for 
38BK76-Locus 3. Although it seemed possible that 38BK76 was established 
earlier than the other sites, there was equally little doubt that the sites 
were contemporaneous for at least part of their occupation ranges. Also, the 
mean ceramic date at 38BK76 was based on a sample of 34 sherds and was con- 
sidered questionable for that reason. An additional explanation of the 
differences in mean ceramic dates, it was thought, would be a compounded time 
lag factor (Adams and Gaw 1977) in view of the hypothesized status differen- 
tiation among the sites. The question of contemporaneity of the sites and 
time lag in the ceramics was, therefore, a guiding question during the 
subsequent mitigation phase. 



68 



TABLE 8. Historic Ceramics Recovered from Site 38BK75-Locus A (Testing) 

Site 388K75-Locus A Surface Materials 

Non-local Ceramics 

Creamware 34 7.3% 

Pearl ware 134 28.9% 

Porcelain 7 1.5% 

Whiteware 14 3.0% 

Delft 3 0.7% 

Stoneware 35 7.5% 

Redware 5 1.1% 

Soft Paste Earthenware 4 0.8% 



236 50.8% 



Local Ceramics 



Colono 228 49.2% 

Site 38BK75-Locus A Excavated Materials 

Non-local Ceramics 

Creamware 1 2.5% 

Pearl ware 6 15.0% 

Porcelain . 0.0% 

Whiteware 0.0% 

Delft 0.0% 

Stoneware 2 5.0% 

Redware 1 2.5% 

Soft Paste Earthenware 0.0% 



10 25.0% 



Local Ceramics 



Colono 30 75.0% 



Total Sample 

Non-Local Ceramics 246 48.8% 

Local Ceramics 258 51.2% 



69 



TABLE 9. Historic Ceramics Recovered from Site 38BK76 (Testing) 



Site 388K76 Surface Materials 



Non-local Ceramics 

Creamware 

Pearl ware 

Porcelain 

White ware 

Delft 

Stoneware 

Redware 

Soft Paste Earthenware 




3 
2 
3 



2 



10 



0.0? 
2.4% 
1.6X 

2.4% 

0.0% 
0.0% 
0.0% 
1.6% 

8.1% 



Local Ceramics 



Colono 



114 



91.9% 



Site 38BK76 Excavated Materials 



Non-local Ceramics 



Creamware 
Pearl ware 
Porcelain 
Whiteware 
Delft 
Stoneware 
Redware 
Soft Paste 



Earthenware 



10 
4 
4 

2 
2 

2 



24 



2.3% 
0.9% 
0.9% 
0.0% 
0.5% 
0.5% 
0.0% 
0.5% 

5.6% 



Local Ceramics 



Colono 



408 



94.4% 



Site 38BK76 Total Collection 



Non-local Ceramics 
Local Ceramics 



34 
522 



6.1% 
93.9% 



70 



en 

UJ 

cr 
< 



lOOn 



90- 



80- 



70- 



38BK75B 



38BK76 



uj 60 - 

o 
or 



50 H 
o 
5 
2 40- 

UJ 

u 

30- 



20- 



10- 



LOCAL / NON-LOCAL 
CERAMICS 




Vsss&a 



45 



55 60 65 70 

AVERAGE SITE ELEVATION IN FEET ASL 



75 




LOCAL CERAMICS 
NON-LOCAL CERAMICS 



note: swamp elevation 

±20" TO ± 26' ASL 



FIGURE 19 

Site 38BK75 and Site 38BK76 
Relative Amounts of Nonlocal 
and Local Ceramics 



71 



It was apparent that Sites 38BK75 and 38BK76 were closely linked temporally, 
geographically, and culturally at the end of testing. This linkage was 
probably stronger than that between Loci A and B at 38BK75 (unless otherwise 
noted, 388K75-Locus A will henceforth be referred to as 38BK75, since Locus B 
was not studied further). However, the preliminary analysis of the sites rec- 
ommended for additional investigation presented more potential questions than 
answers. It was hoped that mitigation of these sites would lead to under- 
standing the function and relative socio-economic positions of each site, as 
well as yielding a substantive contribution to our understanding of the ma- 
terial culture and structure of a slave community. 

It was hypothesized that both 38BK75 and 38BK76 represented slave quarters of 
the same plantation and that the difference in percentages of colonoware and 
mean ceramic dates was a reflection of status differences among the slaves 
occupying the two sites. Site 38BK76 was thought to represent the habitation 
of field hands and Site 38BK75 the location of either house slaves, crafts- 
men, or both. It was proposed for mitigation to examine cultural differences 
as expressed in artifact distribution and patterns in more detail than had 
been possible during the testing phase. For example, it was expected that 
more furniture and personal group artifacts (South 1977a) should have been 
found at 388K75 than 38BK76; and house size and quality should have been 
larger and better at 38BK75 than at 38BK76. It was also proposed to study 
the subsistence base and available plantation technologies, such as colono- 
ware at 38BK75 and 38BK76. 

It was proposed to excavate a 40 x 40 foot block around Feature 1 at Site 
38BK75 in order to examine the feature found during testing and find the 
house suspected to be nearby. At 388K76, it was proposed to excavate a 50 x 
50 foot block around Test Unit 3 to examine the structure and any associated 
trash features suspected to be 1n the vicinity. It was recognized that 40 x 
40 and 50 x 50 foot blocks would possibly not be sufficient mitigation steps, 
and a cautionary statement that further excavation might be necessary was 
made. These recommendations were later downgraded in scope by IAS-Atlanta to 
one 10 x 10 foot excavation at 388K75 and one 30 x 30 foot block at 38BK76 by 
IAS-Atlanta before mitigation began. 

Site 388K245 was not tested by SSI, and in fact, its importance had not been 
recognized until it had been badly damaged by excavation of a large borrow 
pit by construction contractors. On the basis of information supplied by 
IAS-Atlanta personnel and brief visits to the site during testing at the 
other sites, recommendations were developed for mitigation at the site. 
These recommendations were that a brick kiln and cellar recognized by 
IAS-Atlanta be excavated, and a large area of road ruts and other features be 
mechanically stripped and mapped since borrow pit activity had already com- 
pletely stripped the topsoil . An hypothesized kitchen to the north of the 
borrow pit was to be left untouched by construction so that no recommenda- 
tions concerning it were made at the time of the mitigation contract 
negotiations. 



72 



It was felt that 38BK245 represented part of the slave quarters and other 
areas of Curriboo Plantation, which was coeval with Yaughan Plantation 
(38BK75 and 38BK76). Since a brief reconnaissance of the site indicated 
large numbers of colonoware sherds and a high probability of structures and 
trash features, 38BK245 presented a control and base of comparison for Sites 
38BK75 and 38BK76, especially in relation to a study of artifact and 
structural patterns. 

Summary and Recommendations 

Before testing began, it was thought that the five sites to be tested were 
mid-eighteenth to early nineteenth century plantations owned by Huguenots 
(38BK73, 38BK75, and 388K76), an eighteenth century tavern (38BK88), and a 
nineteenth century farmstead (38BK91). Based on these conclusions, certain 
guiding questions or objectives were established for the project. Chief 
among them were to study the differences between French and British culture 
and economies on a local and regional level, to study the economic conditons 
and relationships of plantations and other site types such as taverns in the 
area and to study the overall significance of these sites to contemporary 
archaeological questions. 

To these goals, SSI added several others: Was there enough material and 
adequate preservation at the sites to answer such questions? Could South' s 
(1977a) artifact patterns be applied to the material? and Were there any 
structures around which to base South' s pattern analysis? 

It was proposed to conduct surface surveys, metal detector surveys, test 
excavations, and limited historical research to initially address the ques- 
tions posed. Briefly, the conclusions of the testing phase indicated that 
two of the sites were badly disturbed and three deserved further investiga- 
tion. ' Site 388K88 was not a tavern and contained no structural or trash 
features. Site 38BK91 was not a farmstead, but rather a surface brick clamp, 
badly disturbed by pine plantation. Site 388K73 had disappointingly few 
artifacts but appeared to be a special function structure. Later mitigation 
of 388K73, discussed in this chapter, showed that the site included a tree- 
fall and a sparse field scatter of eighteenth to twentieth century debris. 
Sites 38BK75 and 38BK76 did indeed appear to be slave quarters, but they were 
not capable of answering questions of French ethnicity since it became appar- 
ent (later confirmed during mitigation) that the Huguenots in South Carolina 
had become acculturated into the Anglo-American mainstream before the estab- 
lishment of the sites. A sixth site, 388K245, was not tested by this proj- 
ect, but went directly to mitigation on the basis of work conducted by 
IAS-Atlanta (Figure 17). 

Based upon the field results of the testing phase, it was expected that at 
least one structure and associated features would be present at 38BK75, at 
least two structures and associated features would be present at 38BK76, and 
a brick clamp, cellar, plantation outbuildings, and associated features were 
present at 38BK245. The strati graphic integrity of Sites 388K75 and 38BK76 
was expected to be much better than at 38BK245. 



73 



Because the fundamental conclusions on the character of the sites changed as 
a result of testing, new research priorities needed to be established. These 
necessarily revolved around slaves and slave lifeways at Sites 38BK75, 
38BK76, and 38BK245. It was proposed to examine the hypothesized status 
differences between the slaves at 38BK75 and 38BK76 by carefully excavating a 
structure and associated features at each site. Several patterns were ex- 
pected to indicate status differences. Site 38BK75 was expected to have 
larger slave cabins, more furniture associated artifacts, and more non-local 
ceramics than Site 38BK76. It was proposed to examine the subsistence base 
of slaves at all three sites in order to compare this data on nutrition be- 
tween sites and with that at nineteenth century slave sites in Berkeley 
County and coastal Georgia. It was also proposed to examine the technologies 
of simple plantation industries such as blacksmi thing and fishing to compare 
with data from nineteenth century sites and the literature. It was further 
proposed to define colonoware ceramics found at the site and to identify the 
ethnicity of the makers and their manufacturing techniques. As will be seen 
below, these goals could not be met in all cases, and at the conclusion of 
the mitigation fieldwork they were again modified. 



74 



V. CURRIBOO, YAUGHAN, AND THE CORDES FAMILY 



Introduction 



In order to interpret the archaeological data recovered from Sites 38BK75, 
38BK76, and 38BK245, and to relate the interpretation of this data to larger 
historical issues, a series of preliminary questions were posed and answered: 

1. Who owned the plantations? 

2. Who occupied them? 

3. What was the function of the sites? 

4. What was the nature of the occupation? 

5. What crops were raised on the plantations? 

The historical research was conducted with the archaeological investigation 
and was in a position to respond to questions that emerged as the archaeolo- 
gical analysis progressed. Therefore, as the extent of the artifacts relat- 
ing to slave culture became apparent, questions relating to the history of 
slavery assumed major importance. In the meantime, a systematic search of 
repositories uncovered documents describing both plantations, members of the 
family, neighbors and neighboring plantations. The historical research then 
found itself in a position to answer specific questions and to describe the 
economics of the plantation and its impact on the slaves. It was also able 
to reconstruct detailed information about slave life and the relationship 
between the white family and the slave family. 



Procedures 

The questions asked were shaped not only by the requirements of the artifacts 
and the sites but also by the nature of the written evidence as it was uncov- 
ered. In its broadest outlines, the procedures for discovering sources and 
interpreting them followed the analytic/synthetic method. The problem was 
first broken down into its constituent parts, or constituent considerations 
(e.g., who owned the plantation), and then the interpretation of the project 
as a whole emerged as the constituent conclusions were linked to one another. 
In this way, for example, information on the viability of indigo and the de- 
velopment of cotton cultivation had meaning in the interpretation of the size 
and distribution of the black population on the plantation. 

The procedure for identifying primary and secondary sources followed a paral- 
lel plan. General histories of South Carolina, Berkeley County and planta- 
tion agriculture were read to identify topics of special interest: rice, 
indigo, plantation economics, slavery, inheritance strategies and so on. 
More specialized secondary materials were then located. Having isolated the 
key individuals and the general location of the land, the research focused on 
identifying the land and people as fully as possible through primary, mainly 
archival, materials. As many primary documents as possible were located 
describing the Cordes family, the principal owners of the two plantations. 
Other documents that described the St. Stephen's Parish/Pineville district 
were carefully consulted so that reasonable inferences about the two planta- 
tions could be made when direct testimony was lacking. Research in the 



75 



primary and secondary sources informed each other so that historical issues 
directed attention in relevant channels while ongoing archival research 
refined perception of those issues. The relationship between this project 
and the scholarly literature and research in progress elsewhere therefore 
flowed naturally from this approach. 

In addition to the conventional, written methods for accessing primary and 
secondary sources, individuals were also consulted at various points for 
their expertise in the field and their familiarity with the archives in the 
state. These individuals included Dr. George Rogers, Department of History, 
University of South Carolina; Dr. David R. Chesnutt, The Henry Laurens Pa- 
pers, University of South Carolina; Mr. George Terry, Curator, Historical 
Collections, McKissick Museums, University of South Carolina; Mrs. Martha 
Bailey Burns, Librarian, Huguenot Society of South Carolina, Charleston; Mr. 
James Percival Petit, Isle of Palms; Mr. Elias Bull, Charleston; and Mr. 
Lucas Gailliard, Charleston. 

Documents 

Archival sources were located in several repositories and fall into different 
categories. Land records, used to establish chain of title, were found at 
the Register Mesne Conveyance, Charleston; Register Mesne Conveyance, Monck's 
Corner; Assessor's Office, Monck's Corner; and on microfilm at the South Caro- 
lina Department of Archives and History, Columbia. These records go back to 
the colonial period and include royal grants, warrants, conveyances, plats 
and deeds. Additional information on the transfer of land was found in 
wills. 

Wills and inventories of personal property were tremendously useful. They 
supplied detailed information on what- was on a plantation and identified key 
individuals. Inventories, moreover, listed slaves by name and price, which 
gave a clear picture of the slave community. Whether a given slave was a 
prime field hand or possibly a child could occasionally be inferred from the 
price relative to other prices. In the 1930s, many of these records were 
transcribed and typed by the Work Projects Administration. Original manu- 
scripts, typescripts and microfilm of the records are available at several 
locations: Charleston Probate Court, Charleston County Public Library and 
the South Carolina Department of Archives and History. 

The next block of documents concerned individual plantation records including 
ledgers, daybooks and account books. The Account Book of the Estate of John 
Cordes, which recorded Samuel Cordes' administration of John Cordes' estate, 
was a critical document, supplying more than ten years of continuous infor- 
mation on an estate that included Yaughan Plantation. Evidence from this 
source was confirmed by studies of account books belonging to other inhabi- 
tants of St. Stephen's and nearby St. John's, Berkeley. The John Cordes 
Estate Account Book is at the Library of the College of Charleston; other 
similar materials are located at the South Carolina Historical Society, 
Charleston, and the South Carol iniana Library, University of South Carolina. 

Numerous other manuscripts were consulted, including court records, 
Revolutionary War records, marriage settlements, tax returns, federal and 
local censuses and the South Carolina Gazette , which began in 1732. This 
information was used to develop the context and to support inferences. 



76 



Summary of Ownership 

Title searches were conducted for both Yaughan and Curriboo plantations in 
order to ascertain as much as possible the period in which these sites were 
settled and occupied. Although the title data alone did not establish these 
points, information was obtained from deeds, probate and personal records 
which outlined the parameters of occupation. In both cases, the plantations 
were worked out well after the period in which they had been occupied as 
discrete settlements. In the following summary, therefore, the discussion is 
confined to the period in which households were coterminous with economic 
units; complete title information has been presented in Appendix B. Figure 
20 represents time line summaries of the data presented below and in Appendix 
B. 

Yaughan Plantation 

The earliest reference to Yaughan Plantation dates from 1737 (Figure 21). 
The 650 acre tract conveyed by Richard Allen to Edward Thomas, "known by the 
name of Yaughan", was part of a larger, 1200 acre tract. Thomas left this 
property to his son Samuel, who in turn sold it and the adjacent 596 acre 
property to Isaac Cordes in 1742. The two tracts became known as Yaughan 
Plantation, and although it is not clear whether Samuel Thomas developed the 
land, the inventory of Isaac Cordes' estate in August 1745 lists cattle, 
sheep, hogs, horses and some household goods at "Youshan" (Inventory of Isaac 
Cordes, 9 August 1745, Inventories 67:A:329). 

Isaac Cordes left his real estate to his son John, who died in 1756. John 
Cordes left his real estate equally divided between his two sons, John and 
Thomas, who were both children. His brother-in-law and cousin, Samuel 
Cordes, became their guardian in 1756, and Yaughan appears to have functioned 
primarily as an indigo plantation under the daily supervision of a series of 
overseers (Account Book, Estate John Cordes, 1764-1798). John Cordes at- 
tained his majority in 1768, and although local tradition states that he 
inherited all of his father's property by law of primogeniture (Dubose 
n.d.:50-51), John Cordes clearly continued to administer Yaughan Plantation 
on behalf of his brother Thomas (Account Book, Estate of John Cordes, 
1764-1798). In 1773, Thomas Cordes formally accepted his share of his 
father's estate from his uncle, and his brother John confirmed his title by 
deeding him half of their father's estate in what appears to have been a 
straw-man transaction. 

Thomas Cordes had already begun to participate in local parish affairs, and, 
thus, he had probably begun to reside at least part of the year at Yaughan. 
Although he took an active role in the Revolutionary War, his name appears 
periodically in parish records through the 1770s. He married in 1784, and 
records of various neighbors in the post-war years indicate that he bought 
indigo and rice seed, evidently to receive the plantation's prewar functions 
despite changes in the indigo market, which put the American grown commodity 
at a severe disadvantage. He and his family continued to live at the planta- 
tion until his death in 1809. 



77 



1737 



T 



YAUGHAN 

ALLEIN TO THOMAS , 27 JULY 1737 
THOMAS TO THOMAS , 27 FEBRUARY 1740 



-- THOMAS TO I. COROCS , 2< JANUARY 1742 
-- I. COROeS TO J. COROCS , 9 AU4U3T I74S 

-- J. COROCS TO SONS , 3 DECEMBER 1754 



•- J. COROCS TO T. COROCS , K) MAY 1775 



--( T. CORO£S TO WIPE » CHILOREH ) ,1*04 



-- C. COROCS TO 9 CHILDREN , I92B 



--3 COROO 3I3TCRS TO CLARKE, 1434 



■f CLARKE TO TNURSTOM , I JANUARY 1430 

THUWTON TO MONET , I OCTOMCR 1497 
MONCY TO FLATT , 4 JANUARY I9S4 



■" LEOUCUX / FLATT TO 3T0MCY , 3 JANUARY M)9I 



FLATT TO 3 SONS) , 1404 

FLATT TO C044C9MAU. AND COM4C3HAU., 1912 

C044CSMAU. AMO C004C3MALL TO BROWOCR , 1913 



CURRIBOO 

OWNED 8Y THOMAS EL1_ERY,I737 

OWNED SY COROCS , 1749 

OWNED 4Y T. COROCS , 1744, TO S. COROCS 



S. COROCS TO T. COROCS, JR., 1794 
T. COROCS , JR. , TO J. J. COROCS , 1799 



J.J. COROCS TO C. M««9CTM , 1949 
C. MwSCTN TO J.*. TNURSTOM , 1949 



J. TNURSTOM TO R.C.R.R.C., 1994 

J. THURSTON TO H. FAMZCRSUTER , 1942 

+ J. TNURSTOM TO J. WELCH , 1997 



E.R. RICXEN9AKER , 1997 

C.R. RICXEN9AKCR TO A.E. CARTER , 1909 

A.C. CARTER TO LORENZ , 1912 

LORCNZ TO O.M. DAVIS , 1*4 

OAVIS TO J.E. SELL , 19(9 

J.I. 9EUL TO R.S. SELL , 1919 

-- SELL TO STR0N9 , 199* 



J» STRON4 TO COORER , l»40 



•- LIO 0. BROWOER , 1974 
1979 -L IROWOCR TO SROWDER , 1979 



FIGURE 20 

Chains of Title for 

Yaughan and Curriboo 

Plantations 



78 




650 Acres 



Edward Thomas from Richard 
Alltin 1737 recorded 1741 




1600 



APPROXIMATE 
SCALE IN FEET 






FIGURE 21 
1737 Plat of 
Yaughan Plantation 



79 



After he died, his widow and children began to live at least part of the year 
in nearby Pineville. His widow left the property divided equally among her 
children when she died in 1826, and ten years later, her three daughters sold 
their interest in Yaughan to Solomon Clarke. Clarke owned substantial real 
estate in St. Stephen's Parish and in adjacent St. John's, Berkeley, and it 
is not likely that he spent a great amount of time at Yaughan. In 1850, he 
sold the property to J.W. Thurston, who began to subdivide the property in 
1857. The separate parcels changed hands several times over the next century 
and tended generally to decrease in size. 

Curriboo Plantation 

The plat affixed to the indenture conveying Yaughan from Richard Allein to 
Edward Thomas indicates land to the southeast of Yaughan described as 
"Curriboo Land: the late John Moore Esq" (Figure 21). According to the 1737 
Allein/Thomas transaction, Thomas Ellery had acquired the land belonging to 
the late John Moore, who appeared to have owned both Yaughan and Curriboo at 
one point. Similarly, Ellery may have obtained all of Moore's estate and 
therefore, he too may have owned both plantations at one point. Ellery died 
the following year and left the vast majority of his real and personal estate 
to his wife Ann. 

When Samuel Thomas sold Yaughan to Isaac Cordes in 1742, the description of 
the 650 acre tract referred to lands southeast of Yaughan "belonging to Isaac 
and Thomas Cordes" (Deed Book R-5 : 187 ) . This is probably Curriboo, which 
Isaac Cordes and his brother Thomas, known as "Colonel Thomas Cordes", had 
acquired from Ellery or his estate. The 1745 inventory of Isaac Cordes' 
personal property lists slaves, livestock and "Sundries at Correboo between 
Coll Thomas Cordes and the Estate" (Inventory of Isaac Cordes, 9 August 1745, 
pp. 328-330). Although earlier references to the ownership of this planta- 
tion were oblique, the Cordes brothers had acquired Curriboo by 1745. Thomas 
Cordes died in 1748 and willed Curriboo, then consisting of 1390 acres, to 
his second son Samuel. Between 1745 and 1748, Thomas Cordes severed the 
relationship between Curriboo and his brother's estate, and his branch of the 
family became the sole occupants and owners of the plantation. 

Samuel Cordes bequeathed Curriboo to his eldest son Thomas (d. 1799) in 1796. 
Thomas Cordes, also known as Thomas Cordes, Jr., to distinguish him from his 
cousin Thomas Cordes (d. 1806) who then resided at Yaughan, willed three 
plantations including Curriboo to his only son James Jamieson Cordes, who had 
been born only one year before his father died in 1799. He appears to have 
been brought up in Charleston primarily, and in 1845, he and his brothers- 
in-law, who were his attorneys, sold the 2255 acre Curriboo Plantation to 
Charles Macbeth. Four years later, Macbeth sold the entire tract to Robert 
Press Smith, who then began to divide the plantation. 

In 1858, Smith sold 1300 acres to the North Eastern Rail Road Company. In 
1862, he conveyed 930 acres, including the dwelling house (Figure 22) to H. 
Panzerbeither. Finally, Smith sold the remaining 30 acres of pine! and to 
Jacob V Welch in 1871. A series of investors purchased portions of the 
plantation over the next decades and although the size of the parcels tended 
to decrease, the tracts did not become as small as those that represented 
constituent elements of neighboring Yaughan Plantation. 



80 



"YAH an" 




Estate Thurston 



o 

v. 
<3 







SOO 1000 

SCALE IN FEET 



FIGURE 22 
1362 Plat of 
Cum'boo Plantation 



81 



Cordes Family 

The Cordes family, which owned Curriboo and Yaughan in addition to extensive 
lands in St. John's, Berkeley, during the eighteenth and early nineteenth 
centuries, came to South Carolina in the late seventeenth century (Figure 
23). Anthony Cordes (1661-1712), the progenitor of the South Carolina fami- 
ly, left France either just prior to the revocation of the Edict of Nantes in 
1685 or just after it. Born to the bourgeoisie of Mazamet, Languedoc, in 
1661, Anthony Cordes was educated a physician (Cordes 1974:86; Richardson 
1942:133). References to Cordes and his wife, Esther Balluet, in Carolina 
date from 1696, and their three elder children; Isaac, Magdalen and Esther, 
were born in the colony prior to the formulation of the Huguenots' petition 
for naturalization believed to have been written in 1695 or 1696 (Thomas 
1887:52). This suggests that Cordes and his wife probably migrated from 
Europe to Carolina between 1685 and 1692. 

Anthony Cordes settled on the western branch of the Cooper River, in what 
became the Parish of St. John, Berkeley, either immediately upon his arrival 
or shortly thereafter. In 1706, he took out a warrant for lands adjoining 
the Watboo Barony line, and in 1709, he obtained a second warrant for 200 
acres in Berkeley County at "three Mile head" (Richardson 1942:134). When he 
died in 1712, Anthony Cordes left his eldest son Isaac "my Plantation that I 
now live on Containing two hundred Acres of land, Together with Sixhundred 
[acres] of land Which I Cause to be measured Joyning the Said two hundred 
Acres of land" (Will of Anthony Cordes, 26 January 1711/12, recorded 22 
February 1711/12, Record of Wills, Vol. 1, 1671-1724, pp. 37-38). The re- 
mainder of his real and personal property, including Indian and black slaves, 
he willed equally divided among his four sons and three daughters (Will of 
Anthony Cordes, 26 January 1711/12, recorded 22 February 1711/12, Record of 
Wills, Vol. 1, 1671-1724, pp. 37-38). Since his wife was not mentioned in 
his will, she is believed to have died prior to 1712. 

His children continued to live in the vicinity of St. John's, Berkeley or In 
nearby rural parishes between the Cooper and the Santee Rivers. All four of 
his sons became planters, and his daughters married planters. Magdalen (b. 
1693?) married, for a second time, Peter Simons, a planter in the Parish of 
St. Thomas and St. Denis, which was .situated on the eastern branch of "the 
Cooper. Esther (16957-1757/8) married Gabriel Marion, a planter in St. Tho- 
mas and St. Denis, and Anne (17037-1772) married Peter Simons' brother Fran- 
cis, also of the Parish of St. Thomas and St. Denis. James Paul Cordes 
(16997-1775) married Peter's sister Elizabeth and took up lands in St. 
John's, Berkeley. Francis Cordes (17017-1743?) married Marianne Porcher of 
St. John's Berkeley and settled a plantation on Wassamassaw Swamp. Thomas 
Cordes (16977-1748/9) married Henrietta Catherine Gendron, whose family was 
influential in the neighboring parish, St. James, Santee, Craven County, and 
Isaac Cordes (16927-1744/5) married first Joan Travours of Barbados and sec- 
ondly ElUnor Coker, also of Barbados (Richardson 1942:135-139). 



CORDES FAMILY TREE 



82 




[MARIANNE \_ 
I PONCHCNy 



FRANCIS 
CORDES 
1701?- 

1743? 




CHARLES 
COROES 



OR. 

WILLIAM 

KEITH 



7^ 

/ ANNE 
"I COROES 



JAMES 
COROES 
JR. 

4. I78» 




/ JUOITH \ 
I BANBURY J 



IE 



JOHN 

CORDES 

I749-I7S8 



J I 



PETER 
COROES 




r ..LT7 



THOMAS 

COROES 

1788 -I78» 



(DR.) 

SAMUEL 

COROES 

1790- 1858 



ZTT 



JOHN 

cordes 

k . !7»* 




EDWARD 
SIMONS 
LUCAS 



/ MARY \ 

— I DAVIS J 

V COROES J 




S. 



J SON 

I __ 



TO SOLOMON CLARKE 



SALE OF YAU6HAN (PARTIAL) 
BY CLARKE - 1850 



SALE OF RIVER SECTION OF 

YAUGHAN BY CLARKE'S EXECUTORS -1852 



o 



MALE 



FEMALE 



r , -MALE NOT LIVING 

I | TO MATURITY, 

ACC0R0IN6 TO RECOROS 



FIGURE 23 

Cordes Geneaology 



CORDES FAMILY TREE 



82 





CHARLES 
COROES 



»KTMOMr 

coroes 
iMi-irn 




[MARIANNE \_ 
I R ONC HER J 



FRANCIS 
COROES 
I TO I T - 

I74J? 






JAMES 
COROES 



^3" 




__t77 
i 

THOMAS | 

COROES | 

l7M-ITSt| 

! 




I MARIANNE] 
I COROES I 




REBECCA 
I JAMIESON 



(DR.) 
SAMUEL 
COROES 
I790-I8SS 





THOMAS 
CORDES 

d. I 799 



PETER 
COROES 





I MART \ JAMIESON 

I LUCAS j~ COROES 

\ J k.lTMf 



JOHN 
HARLOTO* 




FRANCIS 

COROES 

1772- ISSS 



EDWARD 
SIMONS 
LUCAS 



/ MART \ 

— ( OAVIS ] 

V COROES J 



NO HEIRS 




TWINS | , 

I I (DIED IN INFANCT) | 3a " , 

I 1 L. I 



r 1 ^ 

'° N 1 1 




D 
O 



MALE 



FEMALE 



TO SOLOMON CLARKE 



, . -MALE NOT LIVING 

I | TO MATURITY , 

i ' ACC0R0IN6 TO RECOROS 

i i 



SALE OF YAUGHANt PARTIAL) 
BY CLARKE - 1830 



SALE OF RIVER SECTION OF 

YAUGHAN BY CLARKE'S EXECUTORS - 1852 



INDICATES PEOPLE WHO OWNEO PROPERTY 
AT CURRIBOO OR YAUGHAN 



FIGURE 23 

Cordes Geneaology 



83 



Isaac Cordes' marriages suggest the possibility of commercial links beyond 
the colony. Carolina's early development was tied to social and economic 
changes on Barbados, and through the colonial period, the trading nexus 
between the West Indies and Charleston was a significant element in the 
colony's growth (Sirmans 1966:58). Thomas Cordes (16977-1748/9) witnessed 
Thomas Ellery's will. Ellery lived in Charleston and clearly speculated in 
lands and engaged in transactions conducted through the Carolina Coffee House 
in London (Will of Thomas Ellery, 2 October 1738, recorded 7 March 1738/9, 
Record of Wills, Vol. 4, 1736-1740, p. 116). This suggests that Thomas 
Cordes (16977-1748/9) also pursued a varied set of economic enterprises. 
Thus, it is not inconceivable that both Isaac and Thomas (16977-1748/9) 
Cordes were planters with additional, significant economic interests. 

The Cordes brothers' numerous land transactions in the early 1740s in the 
vicinity of English Santee, which became the Parish of St. Stephen in 1754, 
are consistent with patterns in land speculation. The Cordes family was, 
moreover, one of six families in St. John's, Berkeley, that lived in the 
parish throughout the entire colonial period. These elite families achieved 
their status through acquisition of land on which to settle children and 
grandchildren and through carefully planned marriages (Terry 1981:111:19, 21, 
24, 37). English Santee was in the process of settlement in the 1730s and 
1740s. The timing of the purchase of Yaughan and Curriboo lends credence to 
this inference, which, in turn, suggests that both properties were acquired 
initially as speculative property but were kept as future security for the 
family. Both Isaac and Thomas (16977-1748/9) Cordes identified themselves as 
residents of Berkeley County, adding further evidence that properties toward 
the Santee River were ancillary to their primary operations. Although Isaac 
Cordes was vague in his reference to real estate, Thomas Cordes 
(16977-1748/9) detailed the division of his extensive holdings to his three 
sons: 

I give Devise 4 bequeath unto my said Son Thomas 
Cordes. . .all that my Plantation commonly called Whiskin- 
boo which I bought of Peter Allston 4 others containing in 
the whole about three thousand acres of Land. . . situate 
in Craven County. . .As also all that my Plantation con- 
taining about One hundred and Ninety Acres of Land. . .situ- 
ate in Berkeley County. . .butting 4 bounding on the Lands 
of my Brother James [Paul] Cordes, As Also all that my 
Corner Lot of Land seituate in Childsbury Town on Cooper 
River 4 lying appositite to Lloyds Corner Lot. . .1 give 
Devise 4 bequeath unto my said Son Samuel Cordes. . .all 
that my Plantation commonly called Correboo containing 
about One Thousand three hundred 4 ninety acres of 
Land. . .seituate in Craven County. . .As Also all that my 
other Town Lott of Land, seituate in Childsbury Town as 
aforesaid, adjoyning to the Corner Lot herein before De- 
vised my said Son Thomas Cordes. . .1 give Devise and 
Bequeath unto my said Son James Cordes 4 to his heirs 4 
assigns for Ever. . .all that my Plantation where I now 
live [in Berkeley County] consisting of Two Tracts con- 
taining togeather about Eight Hundred Acres of Land. . . . 



84 



(Will of Thomas Cordes [16977-1748/9], 25 April 1748, re- 
corded 21 April 1749, Record of Wills, Vol. 6, 1747-1752, 
pp. 142-143). 

Like Anthony Cordes before him, Thomas Cordes (16977-1748/9) saw several of 
his children married within an expanding kinship nexus and wedded to indivi- 
duals from the area between the Cooper and the Santee. Samuel Cordes (d. 
1796) married a maternal first cousin, Elizabeth Porcher, daughter of Peter 
and Charlotte Marianne Gendron Porcher. His sister Elizabeth (d. 1783) mar- 
ried another first cousin, Elizabeth's brother Peter Porcher, who eventually 
settled Peru Plantation in St. Stephen's Parish. Thomas' (16977-1748/9) 
eldest son, Thomas (17287-1762/3), married outside the kinship circle, but 
within the county; he married Anne Ravenel of the powerful Berkeley County 
family. Thomas' (16077-1748/9) eldest daughter, Catharine (1724-1805), mar- 
ried her first cousin, John Cordes (1718-1756), who was Isaac Cordes' only 
surviving son. (A second son, Charles, apparently predeceased his father.) 

In what was most likely another arranged marriage, Isaac's elder daughter 
Mary married Daniel Huger, whose family was prominent in St. James, Santee. 
Isaac's other daughters, Esther and Anne, were apparently single at their 
father's death; Anne married a Dr. William Keith, and Esther married first 
the Reverend Daniel Dwight and then James Keith (Richardson 1942:135, 
140-141). Thus, through judicious marriages, the Cordes family consolidated 
and expanded its control of lands in Berkeley County and began to repeat the 
process in their newly settled lands in the neighboring Parish of St. James, 
Santee, which then included the future Parish of St. Stephen. 

John Cordes (1718-1756) died less than 15 years after his father's death. 
His sons were children at his- death; Charles died prior to 1756, since he is 
not mentioned in his father's will, but John (1749-1798) and Thomas 
(1753-1806) lived to maturity. John Cordes' principal residence was clearly 
in St. John's, Berkeley, where he willed his wife Catharine both her resi- 
dence and "the Liberty of Planting" during her widowhood (Will of John 
Cordes, recorded 3 December 1756, Record of Wills, Vol. 7, 1752-1756, p. 
582). He divided his real estate equally between his sons but stipulated 
that his personal property including slaves, harvested crops and livestock 
was to be divided equally among his children, after his wife's one-sixth was 
withdrawn (Will of John Cordes, recorded 3 December 1756, Record of Wills, 
Vol. 7, 1752-1756, pp. 582-583). He named Thomas (17287-1762/63), Samuel (d. 
1796) and James Cordes, his cousins and brothers-in-law, among his executors 
whom he requested to maintain his estate in order to support his family. 
Thomas Cordes (17287-1762/3), the eldest, apparently took charge of the 
estate at John's death; records surviving of Samuel Cordes' administration 
begin in 1764, which is within one year of Thomas Cordes' (17287-1762/3) 
death (Will of Thomas Cordes, 22 May 1762, recorded 6 July 1763, Record of 
Wills, Vol. 10, Book B, 1760-1767, pp. 450-452). 



85 



Samuel Cordes (d. 1796) was himself a successful planter, with several hold- 
ings in St. John's, Berkeley and in St. Stephen's, and on the north side of 
the Santee River. He was prominent in parish politics, serving on the vestry 
of St. Stephen's in 1754, when it was incorporated (Misenhelter 1977:1, 6). 
Because his two brothers died without children, he fell heir to their 
extensive properties. Consequently, when he died in 1796, his estate 
included five separate plantations and 408 slaves; his personal property was 
appraised at k 22778. Recent estimates of wealth distribution for South 
Carolina at the conclusion of the colonial period put the mean value for the 
entire colony in the period 1757-1762 at k 6039 and the mean value for St. 
John's, Berkeley, at k 5911. Planters in the colony averaged b 17492, and 
clearly Samuel Cordes was among the wealthier individuals in the state, even 
in this category (Bentley 1977:76-77). 

Neither John (1749-1798) nor Thomas (1753-1806) Cordes achieved the same de- 
gree of success that their uncle did. John received his share of their joint 
inheritance in 1768, and in 1774, he leased Yaughan Plantation, among other 
tracts, to his younger brother Thomas (1753-1806). The following year, "for 
and in Consideration of the natural love and affection which he hath for his 
Brother the said Thomas Cordes and also for and in consideration of the sum 
of Ten shillings Lawful current Money of the Province," John Cordes 
(1749-1798) conveyed Yaughan and two other tracts to Thomas (1753-1806), in 
the exchange discussed above (John Cordes to Thomas Cordes, Release of Three 
Tracts of Land, 10 May 1775, Deed Book, R-5, p. 194, RMC). 

John Cordes (1749-1798) married twice; first, to Judith Banbury and then to 
Catherine Marianne Mazyck, whose father, Stephen Mazyck, was a prominent 
planter in St. John's, Berkeley, and whose mother, Susanna Ravenel , was 
another member of the influential Ravenel family. John Cordes (1749-1798) 
was active in revolutionary politics, serving as a receiver of flour and rice 
and as a member of the House of Representatives for St. John's. He later 
purchased Peru Plantation and moved to St. Stephen's Parish (Richardson 
1942:149-150). 

Thomas Cordes (1753-1806) was very young when his father died, and the ac- 
counts that his uncle kept document in surprising detail his early education. 
He was educated largely in Charleston and given the training of a young gen- 
tleman, complete with French and dancing lessons (John Cordes Estate, Account 
Book 1764-1798, pp. 44, 53, CO. He was clearly on the verge of pursuing the 
elite life of a planter when the Revolutionary War intervened. According to 
family tradition, he narrowly escaped hanging by asking time to smoke one 
last pipe. In the interim, his brother-in-law Theodore Gaillard obtained a 
pardon for him from Lord Cornwallis. Thomas Cordes (1753-1806) was probably 
associated with Francis Marion's brigade. Marion was a cousin, who later set- 
tled Belle Isle Plantation in Sto Stephen's (Richardson 1942:144-145), and 
was supplied during the war at various of the Cordes' plantations. In 1781, 
Marion's brigade made camp on one of James Cordes' plantations and purchased 
from him beef, pork, peas, corn and potatoes (AA 1481, Records of the Comp- 
troller General, pp. 1RR-17RR, SCDAH), forage for the horses and rice. In 
1782, Samuel Cordes supplied hogs, cattle and horses to Marion's troop (AA 
1482, Records of the Comptroller General, pp. 1TT-8TT, SCDAH). Additionally, 



86 



both Samuel and Anne Cordes (Thomas Cordes' [17287-1762/3] widow) loaned 
money to the government to support the war effort (AA 1480, Records of the 
Comptroller General, pp. lqq-4qq). After the war, Thomas Cordes (1753-1806) 
was a member of the House of Representatives and participated in the state 
Constitutional Convention in 1790. He married Charlotte Evance of Charleston 
in 1784 and resided at Yaughan in St. Stephen's Parish until his death in 
1806. 

By the late eighteenth century, St. Stephen's Parish was virtually a matrix 
of related families (Friedlander 1979:285-286). At the turn of the century, 
however, the old colonial families began to disappear through the consoli- 
dation of lands, which closed out families, and because of movement of later 
generations elsewhere. Samuel Cordes (d. 1796) left several plantations in 
St. John's and St. Stephen's Parishes as well as in other parishes to his two 
sons, Thomas and Francis (1772-1855). Both of his sons became planters. 
Thomas died in 1799, leaving one son and one daughter. His brother, Francis 
Cordes married and had one daughter, Mary, who married into the Lucas family. 
Her first cousin, James Jamieson Cordes (b. 1798?), the son of Thomas (d. 
1799) and Rebecca Jamieson Cordes, also married into the Lucas family, thus 
perpetuating the custom of marrying children into a limited number of fami- 
lies. Yet the land had ceased to be the most significant determinant in 
forging family alliances. Thomas Cordes (d. 1799) had married into the Ja- 
mieson family, which was a mercantile family that had had a standing economic 
relationship with the Cordes. James Jamieson Cordes left South Carolina to 
join his wife's family in England in 1823 in order to pursue a rice process- 
ing enterprise and then to build a nail factory (J. J. Lucas to Robert Wilson 
1905, p. 92; Lucas Family Papers, 1792-1796, SCHS). The sale of Curriboo in 
1845 was part of a process in which he first turned property in South Caroli- 
na over to an attorney and then divested himself of it entirely. His sister 
Elizabeth married John Harleston in 1819. Harleston interested himself in 
running Cordes family enterprises as well as those of his own family, and was 
a party to the sale of Curriboo in 1845 (Title, 10 June 1845, Deed Book V-ll , 
p. 45, RMC). 

Yaughan Plantation left the Cordes family under somewhat different circum- 
stances. Thomas Cordes (1753-1806) and Charlotte Evance Cordes had seven 
children, five of whom lived to adulthood. Thomas Cordes (1753-1806) en- 
countered hard times and was sued for bad debt in 1788 (George Bedon v. 
Thomas Cordes, Senior, 18 February 1788, Court of Common Pleas, Judgment 
Rolls, 1791-1910, SCDAH). Between 1790 and 1798 he sold off 24 slaves, which 
implies a decline in the viability of his operations (Thomas Cordes to Timo- 
thy Ford and William Henry DeSaussure, 11 March 1790, Miscellaneous Records, 
Vol. ZZ, p. 137; Thomas Cordes to Catharine Cordes [1798], Miscellaneous Rec- 
ords, Vol. LLL, p. 40, SCDAH). In 1800, he sold 15 slaves to his sister-in- 
law Margaret Cantey, who held them in trust for his wife, Charlotte (Miscel- 
laneous Records, Vol. 000, pp. 270-272, SCDAH). 

After his death in 1806, Thomas Cordes' widow and five children apparently 
went to live in Pineville, at least part of the year. Charlotte Cordes was 
assessed taxes on a lot in Pineville in 1824 (Return of Charlotte Cordes, 17 
March 1825, Comptroller General, Tax Returns, St. Stephen's, 1824, SCDAH). 



87 



In her will, she left "the use of my house and Lott in Pineville as a home" 
to her three daughters: Margaret Catharine, Anna Camilla and Charlotte La- 
venia (Will of Charlotte Cordes, 12 June 1826, recorded 27 February 1827, 
Record of Wills, Vol. 37, 1826-1834, p. 240). In 1832, a "Miss Cordes" was 
listed as a head of household in a census of the town taken that year (Census 
of Pineville, 1832, Thomas Porcher Ravenel Papers, SCHS). Samuel Cordes 
(1790-1858) was a member of the Pineville Police Association in 1824, an 
agency organized in response to the threat of slave rebellion ignited by the 
Denmark Veysey scare (Pineville Police Association, Records, 1823-1843, 
SCHS), but he ultimately moved to St. James, Santee. When he died in 1858, 
he described himself as a "Planter also Physician of St. James, Santee" (Will 
of Samuel Cordes, 24 March 1841, recorded 19 August 1858, Record of Wills, 
Vol. 48, 1856-1862, p. 162). Thomas Evance Cordes (b. 1797), the fifth child 
of Thomas (1753-1806) and Charlotte Evance Cordes, was briefly married but 
evidently died without heirs shortly after his mother's death in 1826. His 
estate was assessed at 1585 acres in 1825, but he was taxed for only 10 
slaves (Return of Thomas E. Cordes, 17 March 1825, Comptroller General, Tax 
Returns, St. Stephen's, 1824, SCDAH). Either he leased the land and lived 
from the rents, or he leased slaves and worked the land himself, since 10 
slaves would have been inadequate to work 1000 acres. 

Thomas Evance Cordes had died by 1840, since the federal census that year 
shows three white women in a household headed by M. C. Cordes (presumably 
Margaret Catharine) in the Parish of St. Stephen between the ages of 40 and 
50. The household at that point included 15 slaves ranging in age from under 
10 to over 55 (U.S. Census Office, Sixth Census, 1840, Enumerators' Manu- 
script Reports, South Carolina, microfilm). Charlotte Lavenia married 
Charles Burnham Cochran in 1842, and the federal census of 1850, which enu- 
merated individuals separately, showed three women in the Cordes household: 
M.. Catherine Cordes, aged 53, the head of household who reported $700 worth 
of real estate (evidently the house in Pineville); A. Camilla Cordes, aged 
48; and A. Carlisle Key, a white woman aged 60 (U.S. Census Office, Seventh 
Census, 1850, Enumerators' Manuscript Reports, South Carolina, microfilm). 
Ann Carlisle Key, evidently an old friend of the family, had been left six 
shares in the state bank by Charlotte Cordes when she died, and had come to 
live with Charlotte's spinster daughters between 1840 and 1850 (Will of 
Charlotte Cordes, 12 June 1826, recorded 27 February 1827, Record of Wills, 
Vol. 37, 1826-1834, p. 239). 

Between 1814 and 1836, Charlotte Cordes and her daughters sold off 46 slaves, 
and in 1850, Margaret Catherine was assessed for only 18 slaves. She and her 
sisters evidently retained enough slaves to run the household and probably to 
rent out to neighbors for income, but they sold a large block of slaves to 
Solomon Clarke in 1836, the same year that they sold him their shares in 
Yaughan (M. Catharine Cordes to Solomon Clarke, 1836, Miscellaneous Records, 
Vol. 5Q, p. 239, SCDAH; Title, 1 January 1836, Deed Book M-10, pp. 221-223, 
RMC). It is likely that they also supported themselves from capital invest- 
ments. In 1836, Charlotte Lavenia Cordes held a mortgage from William Cain 
of St. John's, Berkeley (Mortgage, 1 January 1836, Miscellaneous Records, 
Vol. 5R, p. 147). In the same year, Catharine Cordes held Solomon Clarke's 
note for $13,300 (Mortgage, recorded 28 March 1854, Miscellaneous Records, 
Vol. 5R, p. 146). 



88 



Solomon Clarke had been born in St. Stephen's Parish, where he had extensive 
holdings. In the U.S. Census of 1840, he possessed 166 slaves (U.S. Census 
Office, Sixth Census, 1840, Enumerators' Manuscript Reports, South Carolina, 
microfilm). Ten years later, he reported real estate valued at $40,000. By 
that time, he had moved to St. John's, Berkeley where he died (U.S. Census 
Office, Seventh Census, 1850, Enumerators' Manuscript Reports, South Caro- 
lina, microfilm; Will of Solomon Clarke, 26 October 1850, recorded 11 Novem- 
ber 1850, Record of Wills, Vol. 45, Book A, 1845-1851, p. 771). In his will, 
he directed his executors to settle his debts before proceeding with the divi- 
sion of his estate. The sale of Yaughan in January 1850, which had become of 
marginal importance as the move from St. Stephen's to St. John's, Berkeley 
attests, may have been part of the decision to abandon St. Stephen's in an 
effort to clear his estate from debt. The sale of the river section of 
Yaughan in 1852 by his executors was explicitly done to facilitate settling 
the estate (Title, 9 March 1852, Deed Book V-12, p. 619, RMC). 

The tracts changed hands many times over the succeeding years, and the names 
lingered to designate acreage rather than to signify a family's resi- 
dence/plantation. "Plantation" in the sense that it meant a household that 
was coterminous with a unit of production ceased to describe these planta- 
tions when the members of the Cordes family moved away in the first half of 
the nineteenth century. 



89 



VI. MITIGATION METHODS 

Introduction 

The mitigation phase of the historic sites at Cooper River began on May 14, 
1979 and ended on October 19, 1979, for a total of 23 weeks. Until the week 
of August 13, the crew consisted of the Field Director, Thomas R. Wheaton, 
Jr., his assistant, Christine Johnson, and three crew members. The washing 
laboratory kept a sixth person occupied. Between August 13 and September 14, 
the crew was expanded to 14 members in the field to provide sufficient man- 
power to complete the large stripping and mapping operations at Sites 38BK75 
and 38BK76. Again, during the week of October 8, 1979, the crew was expanded 
to handle the emergency work conducted at a brick structure at Site 38BK245. 

Archaeological Field Methods 

Archaeological mitigation of the four sites in question can be broken down 
into several tasks: controlled surface collections, hand excavated blocks, 
machine stripping, and large scale mapping. Controlled surface collections 
at Site 38BK75 was completed during the testing phase. At Sites 38BK76 and 
38BK245 controlled surface collections were made during the mitigation phase. 
Hand excavated blocks were employed at all three sites where such a technique 
would provide the most data, most efficiently. In conjunction with hand 
excavation, machine stripping of large areas was also conducted. As part of 
the stripping operations, all features were mapped and all trash features 
either profiled and completely excavated or profiled and only half excavated. 

The surface collection methods used at Sites 38BK76 and 38BK245 differed. 
Site 38BK76 had not been significantly impacted by agriculture or logging 
activity, except where logging roads cut into subsoil. Before controlled 
collections could be made, it was necessary to clear the site of trees and 
piles of brush and branches left by the most recent logging (Figure 24). 
This was accomplished by a bulldozer, which was also used to uproot and clear 
trees smaller than eight to ten inches in diameter. The action of the bull- 
dozer treads also aided in breaking up the root mat. After clearing with the 
bulldozer, a garden tractor was used by an archaeologist to disk the entire 
site, breaking up the root mat more completely and turning over the soil to a 
depth of one to two inches. Three separate surface collections were then 
made, the first immediately after disking, the second after dampening the 
site using a water truck, and lastly after a heavy rain. The first collec- 
tion produced the greatest number of artifacts, and successive collections 
produced fewer artifacts. 

The sample size decided upon in consultation with IAS Atlanta was 10 percent. 
It was further decided to set up a stratified random sample strategy whereby 
100 foot square quadrats were divided into one hundred 10 foot by 10 foot 
units and a random sample of ten 10 foot x 10 foot units was chosen in each 
100 foot quadrat (Figure 25). This assured a random sample while dispersing 
the 10 foot by 10 foot units across the site (Figure 24). Artifacts within a 
100 foot quadrat, but not in one of the selected 10 x 10 foot units, were 
collected and designated as coming from the general collection of the 100 
foot quadrant. In retrospect, it seems evident that 100 percent of the site 
could and should have been collected using the 10 x 10 foot units. 




90 
SITE 38 8K76, LOOKING EAST, 
MITIGATION PHASE 
SURFACE COLLECTION 




SITE 388K76, LOOKING NORTH WE! 
AFTER MECHANICAL STRIPPING 




SITE38 8K245, LOOKING EAST, 
BEFORE ARCHAEOLOGICAL 
MECHANICAL STRIPPING 



FIGURE 24 

Photos of Field Methods 

at Sites 38BK76 and Site 38BK245 



91 




92 



The method used in the field to locate the 10 x 10 foot units was relatively 
simple. A 100 foot grid was surveyed using a transit, and stakes were placed 
at the corners. To locate the 10 x 10 foot units within the quadrats tapes 
were stretched along parallel sides of the quadrats and a third tape was 
stretched between and perpendicular to them. Using a map with the randomly 
selected units shown on it as a guide, the third tape was moved across the 
quadrat. As one side of a designated unit was reached, its corners were 
measured out from one of the two parallel tapes along the third tape and 
flags were placed at the corners thus measured. This method was quick and 
efficient. The flagged units were then intensively and completely collected. 

It was impossible to use those data to compare collecting success under dry, 
water truck wetting, and rainfall wetting conditions as had been hoped, since 
some of the bags from the three separate collections within individual units 
were accidentally mixed. All surface collection figures represent the total 
of all three collections in each 10 x 10 foot unit. 

The surface conditions of Site 38BK245 differed greatly from Site 38BK76. 
Site 38BK245 had been heavily impacted by stripping of the topsoil before 
mitigation began, and as a result very little horizontal integrity remained. 
The collection strategy was adapted to these differing circumstances with the 
expectation that artifact patterning would not be as useful as it was ex- 
pected to be at 38BK75 and 38BK76. The collection strategy at 38BK245 in- 
volved setting up a 50 x 50 foot grid and intensively collecting all arti- 
facts within each 50 x 50 foot square. The result of this effort was to 
provide whole site artifact totals for comparison with other slave quarter 
sites. 

Hand Excavations 

As noted above, all three sites included hand excavated blocks as part of the 
mitigation effort. Site 38BK75 had the equivalent of a little over a 40 x 40 
foot block; 38BK76 had two 30 x 30 foot blocks; and 38BK245 had one 30 x 30 
foot block. A 30 x 30 foot block was usually sufficient to completely 
excavate a house plus associated features. The blocks were excavated in ten 
foot units. 

Placement of the blocks depended upon surface indications and the results of 
testing. A feature found during testing at 38BK75 was to be excavated. 
House posts found during testing at 38BK76 were to be followed to expose the 
ground plan; and at 38BK245, a surface brick feature and hypothesized 
structure were examined. 

Essentially, the same set of excavation and recording techniques were used at 
all four sites. Upon laying out a grid, the units were opened leaving six 
inch baulks (adjacent six inch baulks resulted in 12 inch baulks between exca- 
vated units). Natural stratigraphy was used throughout the general unit exca- 
vation. As each level was completed, elevations, black and white and color 
photographs, and measured drawings were made, and a square level form was 
filled out. All soil, except sterile red clay subsoil and approximately half 
of the root mat at 38BK76 was dry screened through 1/4-inch mesh. Soil sam- 
ples containing approximately two gallons of soil were taken at each level 
from representative units across the block. The total number of soil samples 



93 



taken from excavation levels and features was 71 at 38BK75, 212 at 38BK76, 
and 177 at 38BK245. Together, these samples weighed over 3,235 pounds (1.6 
tons) after drying. 

After all units selected for excavation were excavated, the baulk profiles 
were drawn, the baulks removed, and the entire floor drawn and photographed. 
Units were usually considered completed at about one to two inches above the 
red clay subsoil, once structural features had become clearly outlined and 
the light sand soil horizon above the subsoil ceased to yield artifacts. 

After drawing the floor, the trash and other large features in the units were 
profiled and excavated in two halves and in one to several natural or arti- 
ficial levels. Postholes were excavated and post shape noted. Postmold out- 
lines did not usually become apparent until two or three inches had been care- 
fully excavated from the postholes. In almost all cases, a postmold was found 
upon excavation. Plans and profiles were drawn of each feature excavated, ex- 
cept postholes which were excavated and drawn in plan only. Soil from post- 
holes was kept separate from postmolds. Photographs were taken and feature 
forms filled out on each excavated feature. Final photographs were again 
taken of the entire block after it had been cleaned and the features removed. 

Records of photographs, soil samples and artifact bags by unit, level, and 
feature were maintained in separate notebooks. Square level forms, feature 
forms, and maps were maintained in separate files. 

Mechanical Stripping 

Sites 38BK75 and 38BK76 were mechanically stripped of topsoil to approximate- 
ly one to two inches above subsoil to expose features. This was accomplished 
after the surface collections and block excavations indicated the horizontal 
and vertical extent of the sites. Site 38BK245 was mechanically cleared and 
then shovel shaved of a thin layer of water laid soil and rills left by pre- 
vious stripping. Machinery was rented from construction contractors working 
on other portions of the canal, and the scheduling of equipment and skill of 
the operators shifted constantly, although generally speaking, once the 
operators understood their task, they did their job well. 

At 38BK75, a bottom loading pan was used to strip off the area immediately 
around and to the north of the excavated block. This machine offered the 
advantage of removing the soil without digging in or spinning its tires, and 
it did not leave rills. At 38BK76 a motor grader was used since this machine 
was more maneuverable (necessary to avoid the stumps). At 38BK245 it was the 
only piece of equipment available. The disadvantages of the motor grader 
were that it frequently spun its tires if too much soil was removed at one 
time, and it had a tendency to leave rills. After the major portion of strip- 
ping was completed, a small garden tractor with a draw blade was used to 
clean up loose dirt and aid in shovel shaving. All areas mechanically 
stripped were then shovel shaved by hand to fully expose features and provide 
a clear, even surface. Artifacts collected during stripping were collected 
and added to the general artifact collection for the site, although such 
artifacts were not be used for intra- site or inter- site analysis. 



94 



Various methods were used for mapping during the project, from direct transit 
readings of bearing, distance, and elevation to perpendicular measurements 
from unit walls and stakes. However, the most extensively used mapping tech- 
nique was triangulation. After baulks were cleared in blocks and after por- 
tions of stripped areas were shovel shaved, any missing grid stakes were 
replaced with the transit. In areas of high feature density, intermediate 
gutter spikes at 25 foot to 50 foot intervals were sighted in. Using tape 
measures, all features were mapped by pulling from two known points on the 
grid. A crew of three, two tape pullers and one draftsman, were thereby able 
to rapidly and accurately map large areas of high feature density. All parts 
of features were mapped by triangulation except for postholes, where the 
center was triangulated and the dimensions measured. This resulted in many 
sectional maps of each site at a scale of 1 inch equals 5 feet. A better 
scale might have been 1 inch = 2 feet or 1 inch = 1 foot, but the resultant 
maps would have been too cumbersome to handle. 

After an area of a site was mapped, all of the major features, excluding most 
postholes and house trenches, were wholly or partially excavated depending on 
time constraints. Drawings, photographs, and soil samples were taken of each 
feature. Auger boring soil samples were taken from house trenches that were 
not excavated. These have not proved to be very useful as interpretative 
data sources, however. 

Large features, including the possible floor scatter at 38BK75 (Feature F5) 
and a cellar and brick clamp at 38BK245 (Structure 245A and Structure 245K), 
were approached diferently. Feature F5 at 38BK75 appeared to be a floor 
scatter overlying two structures which were almost exactly superimposed. 
This floor was divided into two foot squares and each square was excavated 
separately. A second possible floor or cellar scatter at 38BK75, Feature 
F33, was trenched and profiles were drawn. Complete excavation of Feature F33 
was impossible within the time limits available. The cellar at Structure 
245A was excavated in four quadrants. Complete north-south and east-west 
profiles were drawn and all artifacts on the cellar floor were thoroughly 
mapped and removed. The brick clamp at Structure 245K was excavated by 
trenching through its center and then by excavation in quadrants. 

One operation that should also be mentioned here was the emergency mapping of 
part of a trench house found during road widening between Sites 38BK75 and 
38BK76. Within an hour after being exposed, the trench house and several 
associated postholes were covered with two feet of heavily compacted red 
clay. For this reason, the features could only be mapped and not excavated. 

Essentially, then, the field mitigation steps moved from surface collections, 
through block excavations to mechanical stripping and mapping with special 
provisions for important features. 



95 



Laboratory Methods 

During the course of fieldwork, artifacts from each day's work were brought 
to the mobile laboratory in St. Stephen. Once there, they were logged in and 
checked against field records. Washing and a very preliminary assessment of 
the artifacts were conducted. The laboratory director, David Babson, was in 
charge of maintaining records, washing, and several assistants who could oc- 
casionally be spared from the field. Approximately one half of the artifacts 
were washed before being transported to the Marietta laboratory at the end of 
fieldwork. The remainder of the laboratory work was conducted in Marietta 
with a laboratory crew that varied from three to five or more and was direct- 
ed by Mr. Wheaton. 

Before the Marietta laboratory phase could begin, the contract was again re- 
negotiated for the analysis phase. On January 11, 1980, this phase finally 
began, and washing was completed by January 24, 1980. The following cata- 
loguing phase continued until mid-April when the results of the flotation and 
water screening were added to the catalogue sheets. All beads, fauna!, and 
floral material were sent to outside analysts. Other tasks which were sent 
to outside analysts included 70 chemical soil tests and x-ray studies of me- 
tal. The results of these various analyses (except the x-rays) are included 
as appendices to this report. Detailed classification and analysis of colono- 
ware, minimum vessel counts, non-local ceramics, buttons, hoes, and other 
artifacts and features were more or less complete by October 31, 1980 nearly 
a year following completion of fieldwork. 

Essential to the handling of over 30,000 artifacts from the hundreds of 
proveniences developed by the project was the use of a computer. Extensive 
statistical analyses of the material was not attempted as this was not part 
of the contract, and nearly complete mapping and excavation precluded the 
need for extensive statistical analyses as might be required when only 10 or 
25 percent of a site is sampled. The computer was used mainly for data 
management. In order to make the data accessible by the computer, a code was 
developed which incorporated the catalog number and the artifact classifica- 
tion. The artifact codes were developed following South' s (1977a) artifact 
groups, with minor modifications. In some cases, artifacts could not be 
neatly put into one of South 's slots and new slots were provided, but the 
major modification was the Inclusion of colonoware in the kitchen group as in 
Chapter IX. 

For persons using this data in the future, a brief description of the catalog 
code system, is presented here. The catalog code consists of three parts, 
the site or major location, a hyphen, the horizontal placement, a hyphen, and 
the vertical placement. A number such as 75B-U31-2 would mean Block B at Site 
38BK75; excavated Unit number 31; Level 2. Various letter codes which might 
fill the separate slots are: In Part 1 of the code, letters A, B, C, D, etc. 
designate various structures or blocks within a site, and T indicates the ma- 
terial was collected during the testing phase. In Part 2, the letter preced- 
ing the number may be absent, indicating a general site collection, L for a 
controlled surface collection Lot, U or an excavated Unit, F for a Feature, R 
for a Rakeback during testing, B for collected from a Baulk, and A for 
individually mapped Artifacts from the cellar at 38BK245. The third part 



96 



of the catalog code is a number designating the level for codes L, U, R, and 
B. (For code F, the last number indicates the part of the feature collected 
and does not automatically indicate level or quadrant. In order to determine 
the meaning of a number following a feature number code or F code, the acces- 
sion notebook must be consulted). The level number codes in Part 3 are for 
surface, 1 for Level 1 (the root mat at 38BK76), 2 for Level 2 (the dark sand 
layer at 38BK75 and 38BK76), 3 for Level 3 (the light sand layer at 38BK75 
and 38BK76), 4 for Level 4 (the red clay subsoil), and 5 for a mixture of two 
or more levels as resulted from stripping and sometimes from baulks. 

Certain categories of artifacts were noted by their presence and not by 
count. These were not used in the artifact patterns, but were analyzed sepa- 
rately. Included in this group of artifacts were bone, brick, mortar, daub, 
seeds and charcoal . Two other categories were counted but were not used in 
the artifact patterns. These were unidentifiable nonlocal ceramics and un- 
identifiable metal. Seeds were not used in the patterns since differential 
preservation due to soil chemistry would affect bone and seeds more severely 
than other categories of artifacts, and indeed the presence of many seeds may 
not have been the result of human activity at all. Unidentified metal was 
not used in the artifact patterns, since putting it in the activities group 
as done by South would elevate this group out of all proportion as metal 
preservation was exceptionally poor at the sites. Much of this metal would 
undoubtedly have gone into the kitchen and architecture groups, if preserva- 
tion had been better. Bone, brick, mortar, daub and charcoal were not used 
in the patterns since these were not used by South (1977a) in his pattern 
analysis. 

As classification and analysis progressed, various records were maintained. 
These included an accession notebook noting each code number, a description 
of each provenience, the original field bag number, number of bags collected 
for each provenience, the number of soil samples, and any additional com- 
ments. Also for each code, a separate catalog sheet was maintained listing 
all of the artifacts, their quantity, artifact code, any unusual properties, 
the date cataloged, and the initials of the cataloger. For consistency, all 
cataloging on these sheets was conducted by two persons, Lynda Morgan and 
Linda France, in close consultation with Mr. Wheaton. 

Soil samples were floated and water screened in Marietta. A notebook was 
kept and the following data were recorded: accession number, number of bags, 
dry weight in grams, texture, Munsell (1975) color under high intensity 
light, date floated, number of artifacts, presence/absence of faunal remains, 
presence/absence of floral remains, and comments. 

The first step after completely drying all samples was to weigh them and take 
a curation sample of approximately 250 grams. The curation samples were put 
into jars and labelled. Munsell soil colors were taken for each dry sample 
under a constant high intensity lamp. For consistency one person, Maria Almo- 
dovar, took all of the readings. A small sample of soil was crushed and damp- 
ened to check for texture. Ms. Almodovar, herself a ceramicist, classified 
soils into sand (S), clay (C), sandy clay (SO, clayey sand (CS), silt (SL), 
and coarse sand (Cr). While this is often done by soils engineers in lieu of 
grain size analysis and for quick reference, it is not totally objective, of 
course. For our purposes, to maintain some kind of consistency and search 
for possible potting clays, the method worked well. 



97 



Once all soils had been so cataloged, they were floated and screened in tap 
water through 1/16-inch screen. The light fraction was allowed to dry, 
packed in film vials, and sent for floral analysis. The heavy fraction was 
sorted for faunal and floral remains and artifacts. Fauna! and floral materi- 
al was sent for analysis, and artifacts were added to the catalog sheets for 
that provenience. 

The bottled curation samples were then tested for pH with a pH meter which 
had been calibrated against neutral distilled water. As this testing con- 
tinued, 62 curation samples of selected features and soil layers were sent 
for chemical soil testing. 

Three chemical tests were run on the 62 soil samples besides pH. The consul- 
tant for this analysis was Bio-Chem Analysts of Decatur, Georgia. The tests 
were total organic-carbon, total organic nitrogen, and total organic phos- 
phate (PO4). Total organic carbon was derived by the wet-digestion method; 
the total organic nitrogen by the digestion and distillation method; and the 
total organic phosphate by the ignition and direct digestion method. It was 
anticipated that these tests would provide data concerning the function of 
features and be comparable to similar data provided by Drucker and Anthony 
(1979) at Spiers Landing. This did not prove as useful as had been hoped for 
two reasons; not enough samples were provided for detailed analysis of partic- 
ular features and the values obtained at Spiers Landing represent soils which 
appear to be different in chemical makeup from those studied here. This data 
is provided in the discussion of features and in an Appendix C for the use of 
future investigators. 



98 



VII. DESCRIPTIONS OF STRUCTURES AND FEATURES 

Introduction 

Twenty-nine structures were identified during the data recovery phase. The 
structures were found and investigated by hand excavation or mechanical 
stripping. Hand excavation located three complete structures at 38BK75 
(Figure 26), three at 38BK76 (Figure 26), and four at 38BK245 (Figure 27). 
All of the hand-excavated structures at 38BK75 and 38BK76 appeared to be 
domestic, while at 38BK245 the hand excavated structures consisted of a brick 
kiln, an office, a naval stores processing structure, and the cellar of an 
unidentified structure. 

Mechanical stripping exposed a larger number of structures, two at 38BK75, 
ten at 38BK76, and seven at 38BK245. All of the structures found by mechan- 
ical stripping at 38BK75 and 38BK76 appeared to be either domestic structures 
or sheds. The structures found during mechanical stripping at 38BK245 were 
primarily domestic, with the exception of what appeared to be a barn. The 
total number of structures per site was, therefore, five at Site 38BK75, 
thirteen at 38BK76, and eleven at 38BK245. 

The following discussion will briefly describe the structures and their as- 
sociated features, giving limited data by which to compare them. Artifact 
lists are presented with the discussion of the artifacts in Chapter VIII and 
soil data are given in Appendix C. Before discussing the structures and 
associated features, a brief description is given here of the major founda- 
tion construction methods, as foundation type is a significant criteria for 
classifying the structures. Three basic foundation types were found at 
Curriboo and Yaughan; these were trench, posthole, and brick pier. The first 
two had different earthen fill and all three illustrated differences in the 
distance between posts or piers, width, depth, orientation, and general size 
and shape. 

Wall trench construction was the most frequent foundation type found during 
data recovery. The most obvious feature of this type of foundation was a 
long, relatively narrow trench excavated into subsoil. Trenches ranged in 
width from .8 or .9 foot to 1.5. feet, although most were approximately one 
foot wide. The width within trenches varied, and there was more variability 
within trenches at Site 38BK76 than at 38BK245. The trenches were nearly 
vertically sided and flat bottomed in cross section (Figure 28). Depth of 
the trenches was from 1.5 to 2.5 feet below ground surface and usually 
extended a foot into subsoil. Length varied from 9.5 to over 40 feet, 
depending upon the size of the structure. 

Two parallel trenches of nearly equal length usually defined the outline of a 
structure, but in some cases cross trenches were placed midway along the 
structure and at the ends. At 38BK76, some trenches appeared to represent 
additions or replacement walls, parallel to the long side. 




FIGURE 26 

Site 38BK75 and Site 38BK76 

Site Plan Yaughan Plantation 




25 SO 

SCALE IN FEET 



FIGURE 26 

Site 38BK75 and Site 38BK76 

Site Plan Yaughan Plantation 




100 



+ 




+ 



50 100 

SCALE IN FEET 



200 



+ 



+ 



-f 



249C 



^ 



FIGURE 27 
Site 38BK245 
Site Plan 
Curriboo Plantation 




100 




\ + 

\ 



6, 



>»<>. 



. MU 



*S*T 



* ^ 



90 100 200 

SCALE IN FEET 









\ 



RAISED AREA 









\ + 



HYPOTHESIZED OWNERS HOUSE 



/ 
/ 

/ + 

/ 



/ 



* 



+ ! 



+ / 
/ 

/ 



/ 



/ 



y 



/ 
/ 
+ / + 
/ 
/ 
/ 



FIGURE 27 
Site 38BK245 
Site Plan 
Curriboo Plantation 



101 



< 

m 

rat- 
so 
=>o 
xcr 



o 

z 
< 

CO 

< 








u 


C 




c 


O 




a> 


•F" 




s_ 


4J 


CO 


i— 


u 


CM 




aj 




1— 


on 


LU 


<T3 




a. 


U 


on 


^ 


•^ 


U1 


CJ 


Q. 


o 


i— * 


>^ 


s_ 


U_ 


I— 


<_J 




102 



Once a trench was dug by the builders, posts were placed down the centerline 
and the trench was refilled, often with fill different from the surrounding 
soil. The quality of this fill and the average distance between posts pro- 
vided two important clues to the superstructure types. The average distance 
between posts of all trench structures with identifiable posts was 2.2 feet, 
as opposed to the 3.7 feet between posts in posthole foundations. 

The trench fill at 38BK76 consisted of subsoil and minor amounts of topsoil. 
This red clay subsoil fill dries hard and may have been used to make some of 
the colonoware pottery found at 38BK75 and 38BK76. The trench fill at 
38BK245 was strikingly different from the red clay subsoil, being finer, more 
elastic, and often of a gray color. The trench fill at Structure 75B2 was 
sand and barely distinguishable from the surrounding topsoil matrix. At 
38BK245, and to a lesser extent at 38BK76, the clay trench fill had a some- 
what swirled appearance, apparently due to being mixed with water, much like 
mortar. This mixing was far from complete, however. 

Posthole foundations are more familiar on historic and prehistoric sites. 
The postholes at Yaughan and Curriboo had varying sizes and shapes for the 
placement of individual and, occasionally, double posts. The amount of 
variability of post distance in posthole foundations was higher than in 
trench construction. This variability, sometimes ranging up to six or even 
eight feet, possibly indicates a greater variety in the superstructure and 
possibly in the function of posthole structures. As with trench fill, post- 
hole fill varied. In approximately half of the postholes, the fill was 
almost exclusively topsoil. It seems apparent that postholes were filled 
with the same material excavated from them, whereas trenches were intention- 
ally refilled with mixed clay. 

A third and minor foundation type occurred in only one structure at Site 
38BK245, and may have fulfilled a function as an office rather than a slave 
cabin or shed. Structure 245C was built on brick piers which rested on or 
just below the ground surface. The bricks were held together with tabby-like 
mortar and the piers were from five to eight feet apart. This was also the 
only structure with an indoor fireplace and brick chimney. 

How measurements were taken is critical to understanding the data on struc- 
ture size and orientation. The distance between posts (i.e., postmolds) was 
only taken when postmolds were present in trench structures or could be 
placed within .25 foot by extrapolation in posthole structures. For this 
reason, some structures could not be used at all for post distance measure- 
ments (Structures 76C, 76J, 76L, 245F, 245C2* and the house in the dirt road 
at 38BK75), and other structures had only a few measurable postmold distances 
(Structure 245B had four, 76G had four, 76F had three, 76E had four, and all 
the remainder had between five and 19 measurable distances). 

Postmolds varied in size and shape, as did the trenches and postholes where 
they were found. The available data did not indicate any correlation between 
trench or posthole size and postmold size, but it should be noted that only 
four structures had hewn or sawed posts, which were rectangular in cross 
section. These were Structure 245H, an hypothesized barn, where all posts 



103 



were rectangular and charred at the bottom, and Structures 245C2, 245B, and 
245D, where both rectangular and rounded posts were used. The variation in 
post shape and size for unmodified posts within a single structure indicates 
that the builders were undoubtedly using material readily available in the 
surrounding forested areas with little or no preparation prior to use. 

Measurement of the structures to obtain length/width ratios and floor space 
were taken in one of the following ways. When postmolds were distinguishable 
in a posthole, walls were measured from the center of the postmolds. When 
postmolds could not be distinguished, measurements were taken from the center 
of the posthole. The measurements of parallel sides were then averaged to 
obtain the length and width of the structure. Trench structures were mea- 
sured in the same way. 

Orientation, along with location and size, proved to be an important factor 
in settlement pattern analysis and varied significantly within and between 
sites. Orientation was measured by averaging the orientation of the two long 
sides of each structure, clockwise from north. Postmolds were used to mea- 
sure orientation, when present. When only a few or no postmolds were present 
or visible, the orientation was based on a line running through the center of 
the end postmolds and touching all visible postmolds, or by running down the 
center of the trenches. 

Additions such as porches, storage sheds, or extra living space were noted on 
some structures. These were recognized by their usually smaller relative 
size, their location, and their trench or posthole fill, which generally 
differed in color and texture from that of the original structure. Other 
structures exhibited repair work. This was recognized by the form of closely 
spaced or extra pestholes on one of the parallel walls; by the slightly 
skewed alignment of the extra postholes, usually on the outside of the struc- 
ture; by postholes overlapping the sides of postholes and trenches; or by the 
smaller size and often shallower depth of the hypothesized replacement post- 
holes. In a few cases at Site 38BK76, trenches appear to have been replaced, 
or additions added, using trench construction. Since none of these struc- 
tures were in hand excavated blocks, it is impossible to determine whether 
these extra trenches at 38BK76 were replacements or additions. 

The placement of doors and windows is problematic. The "lack of sufficient 
quantities of window glass at any of the structures, clouded by post deposi- 
tional disturbances at 38BK75 and 38BK245, make it impossible to locate 
windows. However, available evidence indicates that some of the structures 
probably had single pane windows. Door location is also difficult to deter- 
mine. This problem is addressed as the structures are described below. 

The only definite chimney was found at Structure 245C, and the only definite 
hearth at Structure 75B, but outside the structure. Possible hearths or ash 
dumps were found at Structure 76A outside the structure and at Structure 76B 
inside the structure. These latter two features were very amorphous and may 
have represented other activities such as soap making (Drucker 1979). There 
were no indications of wattle and daub or stick and mud chimneys at any of 
the structures either by concentrations of fired or unfired clay daub or by 
wattle stains. 



104 



Descriptions 

The descriptions given below are organized from one end of a site to the 
other rather than in the order of excavation or recording. After the 
description of each structure, a discussion of the features associated with 
that structure follow. Features which did not appear to be associated with a 
particular structure are discussed as they appear across the site. The 
discussion of each site also includes tables describing the structures and 
features and the artifact patterns associated with them. Artifacts are 
discussed in more detail in Chapter VIII. The tables describing the struc- 
tures and the artifact patterns are self-explanatory, but the tables des- 
cribing the features require explanation. The function of virtually all the 
features could, in one sense, be described as trash disposal since trash was 
disposed in all of them at one time. "Function" on the tables is a best 
estimate of the primary function of the features. In a few cases, even this 
is questionable and is so designated by a question mark. "Association" is 
always difficult to attribute. This is especially true of features from 
stripped areas. Generally, association means simply proximity. The depths 
given for the features are the maximum depths into red clay subsoil since 
these are critical for determining clay extraction pits and because the depth 
of features in stripped areas could not be measured from the original ground 
surface. The estimated number of artifacts per cubic foot is the least 
precise data category on the tables. These numbers could only be computed 
with any accuracy for regularly shaped features and, in fact, were only 
computed for regularly shaped features which could have potentially been 
primarily trash or clay extraction features. Of course, the figures cannot 
take into consideration artifacts which would not normally be preserved. 
Therefore, it is possible that some features were filled with food wastes, 
but upon excavation contained only a few sherds and a low artifact per cubic 
foot ratio. 

Site 38BK75 

Site 38BK75 was located in a field which had been planted in corn and soy- 
beans before fieldwork (Figures 26 and 29). There is a gentle rise running 
northeast-southwest and dropping from the southwest to the northeast in the 
area of the site. The structures appeared to be aligned with this rise. 
Structures 75B^, and 75B2 were located at the southwestern end of the site 
and were centered on the rise. Structure 75C was located to the northeast, 
and although it was aligned with the rise, it was located on its southeastern 
slope. A final structure was located further to the northeast of Structures 
75B and 75C in a dirt road. Trash and clay extraction features clustered 
around the structures and along the top of the rise. 

Tabular descriptions of the structures and features at 38BK75 are given in 
Tables 10 and 11. Since Structures 758^ and 75B2 were side by side (Figures 
29 and 30), the artifact group data (South 1977a) is actually a mixture of 
these structures. The only feature which was clearly associated with only 
one of these structures was Feature F5, a floor midden associated with 
Structure 75B2« The artifact patterns (Table 12) represent excavated 
material from both structures, including square/level material, postholes, 
and features. 



105 



W 

W 
u. 




Z I 

UJ I 

_J 1 
< 1 




W 1 





LP 

CM CC (O 

cc i— 

UfO c 

=> o) a 

C3 4-> +J 

l. on in 







XL 

CD 
00 

to 



o 



a 
O 



o 
o 



Q 

O 



• o 



in 

CD 
00 

m 




* 



106 




107 



TABLE 10. 38BK75 Structures 







75B X 


75B 2 


75C 


in dirt 
road 


Construction Type 


(1) 


P 


P 


P 


T 


Function 




cabin 


cabin 


shed 


cabin? 


Mean Ceramic Date 


(2) 


1789.8 


1789.8 


unknown 


unknown 


Length 




12.5' 


12.8' 


10' 


+18' 


Width 




11.0' 


10.3' 


10' 


unknown 


Floor Space 




153.0 
sq. ft. 


130.7 
sq. ft. 


100 
sq. ft. 


unknown 


Structural Fill (3) 


ST 


ST 


M 


M 


Recovery Method 




Excav. 


Excav. 


Strip 


Strip 


Orientation 




128° 


128° 


128° 


33° 


Posthole Distance 
Average 
Maximum 
Minimum 




3.23' 
5.10' 
1.90' 


3.17' 
4.30' 
2.20' 


4.50' 
5.80' 
3.40' 


unknown 
unknown 
unknown 



(1) P = posthole, T = trench 

(2) See text for basis of mean ceramic date 

(3) ST = sandy topsoil , M = mixed red clay and topsoil 

(4) Excav. = hand excavation, Strip = mechanical stripping 



The mean ceramic date is based upon artifacts from the excavated units, the 
floor scatter (F5), other features, and excavated postholes from Structure 
75B. This date does not include material from outside the block, from the 
clay extraction/trash features to the north of the block, or from surface 
collections. 



The stratigraphy of the excavated block (Figure 31) at 38BK75 generally held 
true across the stripped areas of the site. Generally, the soil consisted of 
dark organic sand, overlying light yellow sand, overlying red clay subsoil. 
Features began to be defined near the top of the light yellow sand and were 
clearly defined by the bottom of this layer. Plow scars occasionally showed 
up within the light yellow sand. 



108 



at 



ITS 



Lf) 

CO 
CO 
oo 



"O co u 
a» -t-> i— 
*-> o .o 

<T3 <C 3 4-» 
E <*- O O 
i- i- o 
■!-» -M S. U_ 

<s> s- o> 

LU < O. 



at 

Q. O TJ 

o c <_> 



CO 

< 



OL 

its 

00 



u 

o 
CO 

< 






(TJ 1- 

o 

r— co 
i- jO 
O 3 
CO CO 

Q. 

3|> 

a» 
x -a 

•r- <U 

e t. 



'.O 



"° ■— £> 

<TJ O «tJ i™ - 

co O O 

>> Q. O 

US o s» ts 

(— -m <e <u 

■a "o 3 

<u c ••»-> 

'- r— ITS <— ^ f— 

•^ CO •?— 1- 1— 

•o O O "O o 

<u a» -^ co v) 

x a. s- a.T3 .o 

1- O ITS O C 3 

S +j -a +j «j ui 



c 


i— 


C 


r— 


T3 


•r- 


ITS 


•f"" 




o 




o 


i— 


(/) 


,— 


CO 


•r— 


jS 


•r- 


jQ 


o 


3 


o 


3 


co 


(/> 


</» 


CO 


Q. 




a. 




B 


& 


2 


£ 




^— 




r— 


T3 


u 


■a 


U 


O) 




at 




X 


■a 


X 


"O 


►^ 


Ot 


«l— 


<D 


E 


5- 


E 


&. 



o 
CO 

a. 
o 
■•-> 



2 
O 

•r— *~ «t— j2 

O O O 

CO «/» CO J^ 

Q. Q. Q. 5- 

O O O TJ 

+j +J +J -O 



cr> 
o 



Lf> 



LO 



cm 



o 

CM 



LO CM 

~H CD 



LO 

o 



f— t i— i vo 



<rs 

> 
o 



C\J 

•a 



CO 

to 

co 

a> 

i. 

3 
■•-> 

U 



00 



<ts 



<u 

s- 
i. 



CM 

o0 



ca 

LO 

co 

i- 

3 
•M 
O 

3 
i_ 
+-> 
O0 



c 

3 

o 



<ts 



en 

c 
<o 

at 



cm 


C\J 


OS 


<>0 


i-H 


.— I 


CO 


CO 


in 


Lf) 


r*- 


P^ 


</» 


crt 


<u 


O) 


s- 


i- 


3 


3 


■*■> 


•M 


U 


U 


3 


3 


L. 


i- 



oo 



oo 






at 



CM 



CQ 
Lf) 

w> 

s- 

3 

u 

3 
V. 
*-> 
OO 



> 
o 



CM 



CO 

on 
r-* 

to 

at 

t- 

3 
4-» 
O 
3 
S_ 
4-> 
OO 



T3 

C 
3 
O 



IT3 

> 
O 



> 
o 



3 

cr 
to 



CM 


CM 


^H 




CO 


CO 


OO 




LL. 


U_ 


U_ 




o0 


oQ 


«a 


O 


i-* 


o 


o 


to 


OO 


OO 


oo 


f^ 


Ll. 


Ll. 


Ll. 


a) 


w> 


<Sl 


</) 


i_ 


at 


a> 


at 


3 


s_ 


i_ 


i- 


+J 


3 


3 


3 


O 


+-> 


+J 


-4-> 


3 


«S 


<T3 


<o 


L. 


at 


a> 


at 


+J 


u_ 


Ll. 


Ll. 


00 



VJ 




o 




co 


CO 










Ifl 








(Q 


(O 










i_ 










-t-> 










c 




i. 




c 


c 


C 


C 


C 




O 




+* 




o 


o 


O 


o 


O 


o~ 


•^ 


V. 


X 




■fM 


•^ 


•»— 


•^ 


•r— 


i- 


4-> 


at 


at 




+J 


+J 


■M 


+J 


4-> 


<v 


<J 


+j 






u 


u 


O 


a 


(J 


4-> 


<o 


4-> 


>» 




ITS 


ro 


IQ 


<T3 


lO 


■M 


i. 


<o 


<r> 




V. 


i_ 


&. 


s. 


&. 


<o 


4-> 


u 






■M 


■*-> 


4-> 


+J 


+J 


U 


X 


co 


U 




X 


X 


X 


X 


X 


CO 


at 


1_ 




4-> 


at 


a 


at 


0) 


at 


s. 


>> 


o 


co 


S. 


>» 


>» 


>» 


>> 


>* 


o 


A3 


o 


■a 


<« 


its 


(0 


(O 


its 


ITS 


o 


r— 


r^ 


s- 


at 


r— 


r - 


r^ 


r~ 


^ 


t— 



o 



o 



CM 

Ll. 



Lf) 



CM 
CM 



LO 
CM 



CM 

I 

LO 

CM 



CM 
Ll. 



o ^ 
oo oo 
Ll. Ll. 



CM 0O 

OO OO 

Ll. Ll. 



109 



LO 

s»i 

CQ 
CO 
CO 



CU 




u_ 


>> J- 


CQ 


-O 3 


LO 


•M 


r«* 


</l <TJ 




C 0) 




w u_ 




41 




■»-> "O 




+-> c 


cu 




A3 <Q 


o 




a. 


c 




<u 


CU 




w i_ 


*" 




O 3 


c 


LO 


<TJ -M 


cu 


Jj_ 


<«- (J 


> 


CQ 


•»- 3 


o 


to 


-t-> i. 


i- 


r»^ 


i. +-> 


a. 




< (/) 







CM 



—J 
CO 

< 



CSJ 

og 
lo 

CM 

u. 

CO 

IT) =»fc 



IT) 
CVJ 
Ll_ 
CQ 
LO 



a* 



CVJ 
CM 



»« 



*« 



»« 



CM 

CO 

lo 



a« 



CQ 
LO 



O 

i- 



1- 



«G- 


o 




o> 


r*. 




.—1 


en 


o 


T— t 


o 




CM 


LO 




r-» 


CM 


o 


• 


• 


1 


• 


• 


1 


• 


• 


• 


LO 


CM 










f— 4 




o 


r» 


CM 














o 


CO 


r-» 


o 


1— ( 


CM 


o 


LO 


pH 


o 


LO 


r>* 














LO 


CM 
















CO 


— H 


LO 










CO 




o 


<T> 


CO 










r»» 




o 


• 


• 


1 


1 


1 


1 


• 


1 


• 


O 


LO 










CM 




o 


o^ 
















o 


o 


r^ 


o 


o 


o 


o 


CO 


o 


o 


o 
















w— 1 


«— • 
















■— 1 


LO 


er> 












LO 


o 


LO 


<—• 












«— 1 


o 


• 


• 


1 


1 


1 


1 


1 


• 


• 


^" 


**• 












fi 


o 


LO 


«* 














o 

1—1 


f*^ 


CO 


o 


o 


o 


o 


o 


«— t 


LO 


^ 


CO 














CO 


CO 


*!■ 






LO 




r*» 


.—1 


o 


co 


LO 






LO 




o 


*3- 


o 


• 


• 


1 


1 


• 


1 


• 


• 


• 


o 


LO 






«— 1 




CM 




o 


LO 


CO 














o 


LO 


LO 


o 


o 


<? 


o 


LO 


^-1 


CM 


^> 


CO 














*T 


«— 4 
















CM 


*T 


CO 






CO 




CO 


LO 


cn 


00 


r»- 






en 




*fr 


o> 


en 


• 


• 


1 


1 


• 


1 


• 


• 


• 


r»* 


LO 










CO 


r-4 


cn 


LO 


CO 














en 


00 


CO 


o 


o 


CM 


o 


r^ 


*r 


«d- 


1—1 


p» 














o 


»-H 
















CM 


Tf 


CO 


LO 


LO 


00 


00 


r-. 


00 


o 


O 


—4 


o 


«-H 


LO 


o 


^ 


*3- 


o 


00 


CO 










CM 




o 


LO 


CM 














o 


CO 


CO 


CO 


CO 


a\ 


<3- 


CO 


*»■ 


r» 


en 


o 






CM 




CM 


CM 


00 


CO 


*r 










T-H 




cn 


CO 


t— i 

cu 

s. 

3 


cu 








cu 


at 


*r 




u 


s_ 




en 


^~ 




•»" 




c 


<u 


3 




c 


<o 


o 


•M 




<u 


4-> 


■♦-> 




•*• 


c 


u 


T— 




-C 


T- 


•»— 




-C 


o 


u 


> 




o 


.e 


c 


</) 


4-> 


CO 


<o 


*i— 


<o 


-w 


u 


t- 


s 


o 


u 


.a 


<M 


■!-> 


•^ 


i» 


3 


£. 


r— 


CD 


o 


U 


O 


^ 


«c 


u_ 


< 


<_> 


Q- 


H- 


< 


1— 



110 



B« 



CO 
CO 

u. 
lo 





CM 


00 




CM 


r- 


a* 


• 


• 




CM 


r^ 


CM 


r^ 


CM 


co 






Ll. 






lo 






r*» 








LO 


o 


=*S 


CM 


-~l 



3 
C 



O 
U 



CM 



UJ 



co 

< 



<U ro 



> 
o 

a. 



*« 



LO 



** 



o 
co 

lO 



*« 



CM 
CM 
U_ 

CQ 

lo 



O 
i- 

-t-> 
u 



r^ 


«— t 


cm 


CM 


• 


• 


CO 


00 


to 


CM 



WO 

to 



CM 



CO 
CM 



-H O 



O O 



I I 



o o 



en 
lo 



CO 



••o 
to 



CM 






CM 

cm 



en 


CM 


to 


CO 


CO 


CO 


• 


• 


1 



CO CM 

r»« — < 



oo •— • 
00 to 






o -« 



to 



o 

CO 



00 LO 

CM «-l 



—« O 



I I 



o o 



LO 



LO 



o 

CM 



o 

CM 



lO »-H 



CM 

lo 



LO 



LO 
CM 



*3- 


«* 


i— 1 


1** 


LO 


o 


<o 


LO 


CM 


Tj» 


00 


o 


• 


• 


• 


• 

CM 


• 


o 
o 



^ CO 
lO CO 
LO CM 



LO 



lO CM 



CO 
CM 



CO 















</> 






cu 










co 






S- 










a. 






3 










*p— 


(/> 




■*-> 


0) 








a. 


<v 




u 


s_ 




en 


pa 




••■» 


c 


cu 


3 




c 


<TJ 


o 


4-> 


<D 


4-> 


•M 




«p- 


C 


u 


•?■» 


f 


••-» 


•?■■ 




-C 


o 


u 


> 


(-> 


.c 


C 


</> 


*■> 


(/> 


« 


-,— . 


4-> 


o 


i_ 


E 


o 


s- 


-O 


+J 




i» 


3 


i- 


1— 


O) 


o 


o 


^ 


<: 


LL. 


< 


o 


a. 


1— 


< 



CM 
CM 



CM 
CM 



CM 
CO 



O 

o 

• 

o 
o 



LO 
CO 



o 
o 

o 
o 



LO 
LO 
CM 



CM 
CM 



CM 
CM 



er> 



CO 
CO 
CM 



O 



Ill 



(0 

<r 

u. 
O crt 

cot 

-J 3 

g o 
cr 
Q- en 



3 
< 
CD 

r- 

co 

UJ 

5 



in 
to 



* 
to 

to' 
to 

(O 

t- 
Z 
3 



en 

_i 

3 
< 
CD 



CC 

o 






to 



CM 

to* 

m 

<r 
to 

m 
to 



m 



in to 




UJ h- 




=!z 




fe 3 




O 




? u. 




Q. o 




(O 




X 




-I 




3 




< 




CO 




X 




r- 




<r 




o 


CVJ 


z 


<r 




r- 








z 




3 







_vi 








u 








o 








p— 








CO 






tn 








i — 


c 




<— i 


:*. 


o 




ro 


CD 


■pa 


00 




CO 


■M 


<U 


UJ 


ro 


T3 


^~ 


e: 




> 


•t— 


=3 


0) 


<T3 


<«- 


tfl 


+■> 


O 


O 


k— » 


•^» 


X 


s- 


Lu 


to 


UJ 


O- 



11 



5 3 



■=>■• 



****« 




— -o 
r c 

S3 
?1 



s '.-•.-■ 



$ 







STRUCTURE 75 B, 
LOOKING SOUTHWEST 



112 




STRUCTURE 75 B 2 
LOOKING WEST 




STRUCTURE 76 A 
LOOKING WEST 



FIGURE 32 

Photos of Site 38BK75, 

Structures 75B, and 75B ? , 

Site 38BK76, and Structure 76A 



113 



Figures 30 and 31 illustrate the floor plans of the structures in the 
excavated block at 38BK75. Structure 75B2 was parallel to and slightly 
offset from Structure 75Bi, the structure immediately to the northeast. The 
shape and size of the two structures were nearly identical, indicating that 
they were built following the same plan and at approximately the same time. 
As can be seen, Structure 75Bi has two replacement postholes on its 
northeastern side and one replacement posthole on the porch or stoop addition 
on the southeastern end of the structure. Structure 75B2 had one replacement 
posthole on the porch or stoop and an extra post has been added on the 
northwestern ends of the northeast and southwest walls, similar in placement 
to those on the northeastern side of Structure 75Bj° The determination of 
replacement postholes, in this case, took into consideration the overlapping 
of postholes, such as the one on the porch of Structure 75B2» the smaller 
size of the designated replacements and their irregularity of placement. 

The replacement postholes at Structure 75B^ were excavated, as were the other 
postholes. The average depth below surface of the original posts was 1.73 
feet, whereas the replacement posts averaged 1.52 feet. When the stoop 
replacement post was eliminated, the average for replacement posts was 1.39, 
or .34 foot shallower than the original posts, although the range of depths 
of the original posts ran from 2.47 to .98 and, therefore, included the mean 
and range of replacement posts. 

Posthole Features F3 and F4 on Figure 30 were .97 and .70 feet below surface, 
respectively. These were not excavated postholes, but true postmolds. 
Feature F3 also showed evidence of having been sawed or hewn on one side. 
These posts' alignment, depth, and form indicate that they were not 
associated with the structures and were probably recent fence posts. 

A floor scatter (Feature F5) covered Structure 75B2« The floor scatter 
varied in thickness from less than 1 to 3.5 inches in depth and covered the 
area noted in Figure 30. Its surface had been severely disturbed by plowing. 
It covered most of the postholes and trenches of Structure 75B2 and first 
appeared in Units 29, 30, and 43. Since at that time it was thought to be 
another trash feature around Structure 75B ^, and because no time had been 
scheduled for continued excavation at Site 38BK75, the portion of Feature F5 
in Units 29, 30, and 43 was sectioned and excavated, revealing a line of 
postholes. After further negotiations with IAS and the discovery that the 
feature was probably a floor, it was exposed, gridded into two foot squares, 
and excavted by the grid. Upon completion of the excavation the remaining 
postholes of Structure 75B2 appeared, and the faint outlines of sand-filled 
trenches of a third structure were noted, into which the postholes intruded. 
No postmolds were noted in the underlying trench structure, leading to the 
assumption that the posts in the postholes directly replaced posts that had 
been in the trenches. If that was the case, then the superstructures of the 
trench and later posthole structures were probably very similar, if not the 
same. 



114 



Analysis of the floor scatter, Feature F5 (Figures 33 and 34), sheds light on 
the interior layout of the structure and upon the placement of the trefoil- 
shaped replacement posthole in the center of the northwestern wall. Table 12 
shows the occurrence of the Architecture Group artifacts (which together 
represent 93.55 of all artifacts from the floor). These artifacts clustered 
along the northwestern wall and along the central axis of the structure. 
Unfortunately, such a spatial distribution could not be conclusively proven 
since the northern portion of the structure was not excavated in two foot 
squares. However, if the artifacts found in the northern portion were aver- 
aged into 15 hypothetical two foot squares to cover the northern portion, 
each square would average 3.5 artifacts. This would not be enough to signi- 
ficantly change the clustering noted above, and indeed, most of the artifacts 
in the northern portion of the feature would probably cluster along the back 
wall (northwest wall), leaving few if any artifacts for the remaining 
squares. In any case, in the two-thirds to three-fourths of the floor excava- 
ted by two foot squares, there was definite clustering along the back wall 
and down the center of the structure. 

Primary de facto refuse as defined by South (1977:297) tended to cluster in 
unused areas or in areas of poor visibility such as corners, soft earthen 
floors, and beneath cupboards. The small size of the artifacts recovered 
from the floor scatter showed that the artifacts would have been easily 
overlooked, could have fallen into cracks or narrow inaccessible areas, and 
could have been easily pushed into the earthen floor (Fehon and Scholtz 
1978). The fact that these artifacts in Feature F5 were concentrated along 
the back wall and down the center of the structure supports an argument for a 
wall down the center of the structure. 

If a line is drawn, bisecting the central artifact cluster and stopping at 
its southeastern end, two things can be noted (Figure 35). First, the struc- 
ture is bisected into two nearly equal parts and second, the line terminates 
at the trefoil posthole in the back wall. Both of these facts are signifi- 
cant. It is felt that the central cluster of artifacts illustrates the 
presence of a temporary wall or screen as defined by the bisecting line. 
This wall or screen would have been approximately ten feet long and would not 
have reached the front of the structure. This would have allowed passage 
from one room to the other without going outside. The peculiar shape of the 
terminating posthole indicates that it was dug after the structure was built 
and owed its shape to being dug while the existing back wall was in place. 
The lack of a posthole in the center of the back wall at 75B^ may indicate 
that this structure had no central dividing wall. However, the lack of a 
comparative floor scatter at 75B^ makes it impossible to state this with any 
certainty. The main argument against such a central wall is that there was 
no posthole for its southeastern end. This can be explained if the wall was 
temporary and was suspended from above. If such a central wall or screen did 
exist in Structure 75B2, it would imply two sleeping areas, one for parents 
and the other for children, or for two couples or sets of adults. The number 
of persons inhabiting the structure would, therefore, have probably ranged 
from two adults (one in each room) to two or more adults and several 
children. 



115 



J5P 




h» 


uj vu < 




« 


«J> 




O 




X 


U ^ 




t- 


,°uj 


»- 


z 


ui 


o 


Ui 


1<rt , . 


«t 


2 


55* 




UJ 

o 


K 


< 


< 


J 


z a 




X 


O">o 


— 


K 








u 






(0 






4- 






•^~ 






•M 




ID 


i_ 




!_ 


< 




a 


I 




s_ 


r— 




13 


< 




■M 






<o 


U_ 




<JJ 


O 




Lu 


C 




LO 


o 




r^ 


• f— 


m 


i«i 


■4-> 


<-o 


ca 


3 




00 


j2 


UJ 


ro 


•r— 


CSS 




i_ 


=3 


ai 


■f-> 


o 


•M 


00 


1— I 


•f"» 


'^ 


u_ 


tO 


O 



\ 






( 



* • • • *sT\ / 

•••••ey t — U 






3 



o 

u. 
u. 

O UJ 

Eg 

S uj 
-i u. 




o 
o 



z 

u. O 

op 

t- < 
5 ° 



116 




117 




LO 





<u 


fO 




s- 






3 


1— 




+J 


^» 




ro 


(13 




O) 


3 <TM 




U_ 


i — LO 




LP 


ro r^ 




l^» 


<J 


LTl 


^ 


•r- a; 


en 


CO 


•M S_ 




CO 


<1) 3 


LU 


CO 


JZ ■*-> 


ci 




-t-> u 


3 


d) 


O 3 


cs 


<u 


Q. i- 


^-< 


•*- 


>>-l-> 


u_ 


00 


X <*n 



o 

a. 



118 



Feature F25, just to the north of Structure 75Bi, was the only definite 
hearth found during the entire excavation, other than the brick chimney base 
at 38BK245. Figure 36 shows that the northeastern post was replaced and 
moved farther from the fire pit. Because both the floor scatter (Feature F5) 
and the hearth were totally excavated, artifact pattern comparisons can be 
made between two clearly defined features of different function. For all 
classes of artifacts other than the Kitchen and Architecture Groups, the two 
features were very similar. However, the hearth showed a very high percent- 
age of Kitchen Group artifacts and a much lower percentage of Architecture 
Group artifacts than the floor (Table 12). This is to be expected if the 
functions of the two features have been correctly identified. 

The depths and locations of the two features are also indicators of their re- 
spective functions. The floor rested on the postholes and trenches at 
Structure 75B2 anc * na( * a depth of approximately .82 foot below surface. The 
hearth was outside of Structure 75Bi and had a depth of 2.69 feet below 
surface for the fire pit, 2.65 feet for the western post, and 1.65 feet for 
the eastern replacement post. Thus, the floor rested below the topsoil and 
on the red clay layer, whereas the hearth was excavated into the clay layer. 

Feature F2 was located to the west of the structures (Figure 37) and extended 
1.1 feet into red clay subsoil. The main portion of the feature was relative- 
ly steep sided with a scatter feathering out on the edges; this was no doubt 
caused by plowing. There appear to have been two episodes of fill (Figure 
37). At the interface between the two filling episodes, hoes were found 
lying on the surface of the bottom layer. Approximately 90 percent of the 
artifacts came from the uppermost layer and 10 percent from the bottom layer. 
This suggests that the feature was filled with naturally occurring soils to 
begin with and only later intentionally filled with trash. Because of its 
depth into subsoil and depositional history, it is hypothesized that the pit 
was originally dug for clay extraction and later filled in as a trash pit. 

The chemical tests of the fill soil show no significant difference from the 
natural soil matrix (Appendix C). If the feature had been meant for disposal 
of animal wastes (no bone was found), the level of phosphate should have been 
higher than the surrounding soils. It should also be noted that the pH level 
was within a range that should have allowed for some preservation. At 
38BK245, the pH was generally very acid (below 3.5), and yet many bones were 
preserved, albeit in poor condition. 

Feature F22 (Figure 38) may also have been a clay extraction feature although 
it only extended 50 foot into the red clay subsoil. The low number of arti- 
facts again indicated that the feature's function as a trash pit was second- 
ary. Soil chemistry, as at F2, was similar to the surrounding soil matrix 
(Appendix C). 

Features F26 and F27 (Figure 39) were not recognized as separate features 
until the top layers had already been mixed. The feature was first sectioned 
east to west at the approximate center, and the south half was excavated and 
then profiled. Upon excavation of the north half, Feature F26 bottomed out 
and showed that it had been intrusive into F27. At that point, it was too 
late to prevent mixing of the northern part of F27 with all of the F26 materi- 
al . For this reason, the features are described together here. 



119 



E 

a 



w 

CD 



Ui 


2ui 


-1 


UJ-I 


o 


oo 


X 


<x 


(- 


Jt- 


CO 


0.(0 


o 


UJO 


0. 


CI 



LD LP 
I — C\J 

U3 ^ U- 

CO 0) 

UJ OO i_ 

a: 3 
=3 a; «m 

i-. -r- <D 
U-(/)U. 



c 
a 
en 

> 



c 
a 

CO 

c 
5 
o 



o 

a 



c 
« 
o 

c 
o 
a 




u. 



o 
o 



a 

"2 • 
x c 

LU O 
M 

s 3 

o o 

2* 

3 
CO o 

e 
>» — 
a 

o"2 

a 

O Q. 

cr 3 



o 
a. 

•a 

•» 

o 

£ 
o 
u 

o 





















.*..'., 




^ 




•o* 


|s- ; ' 






■ ; ;'.-' 








« 










:' .". •" 




E^ 




« 




£»-■■■> .-.-. 








m ? 



CO 



120 




UJ 



UJ 




PLAN VIEW 




Dark Brown Ashy Sand 



Brick Fragments 



JJB 



Light Brown Sand 
Oark Brown Sand 



i4 



SCALE IN FEET 



GRID NORTH 
PROFILE 



4 FIGURE 37 
Site 38BK75 
Feature F2 



121 





rLn 

I 2 

SCALE IN FEET 



WEST PROFILE 




Mottled Dark Brown 
And Tan Send 






te>"---.-;? :: ". 



Charcoal Concentration 



Hard Packed Red Clay 



FIGURE 38 
Site 38BK75 
Feature F22 



122 




L 



co 

CM 



-a 

c 

<T3 



CM 



LO CM 

r-- u_ 
en i^ 

on as 
cc 

cc 

=> CU 

C3 +J 



c/1 
<U 

3 



<T3 

l— i -r- O) 

U. 00 u_ 







123 



Feature F26-27 extended 1.6 feet into subsoil leading to its hypothesized 
designation as a clay extraction pit ultimately filled with trash. The 
nearly vertical sides of the pit which continued even into subsoil also 
indicated that an attempt had been made to extract as much clay as possible 
given the surface opening. Soil chemistry also supported the hypothesis. 
The soil chemistry of F26-27 fit the pattern of the surrounding soil and gave 
no indication that the feature had contained extra amounts of organics. 

Feature F8 was something of an enigma. On the one hand, it appeared to be 
closely associated with Feature F26-27, and on the other it was filled with 
recently rotted wood fragments and contained no artifacts. The feature ex- 
tended over two feet into the red clay, which was a greater depth than was 
necessary for a fence post. In sum, it was concluded that F8 was not assoc- 
iated with the slave occupation of the site and represented an exceptionally 
deep recent fence post. 

Feature F29 (Figure 40) was excavated 2.0 feet into subsoil, and had a high 
artifact total and higher artifact per cubic foot value (24.7) than the 
previous features. Compared to all other features at 38BK75, this feature 
also had more bone. Feature F29 contained 23.28 grams of poorly preserved 
bone, the only such feature in association with Structures 75B^ and 75B2* 
All but 6.5 grams of this bone was found in Level 1, which extended into a 
lower depression in the east side of the pit. This correlated well with the 
phosphate and pH readings for Level 1, which were significantly higher in 
phosphate and more neutral in pH than the normal soil matrix (Appendix C). 

Despite the higher artifact per cubic foot ratio and the presence of moderate 
amounts of bone, it was concluded that F29 was dug primarily for clay extrac- 
tion and was used only secondarily as a trash disposal pit. This hypothesis 
was supported by a layer of water laid sand lining the bottom of the feature 
(Level 2B), and by the number of artifacts in Level 2 compared with Level 1. 
Level 2, which included the water laid sand, contained 14 percent of the 
feature's artifacts, whereas Level 1 contained 56 percent. Apparently the 
feature was dug, allowed to remain open, and then slowly filled until Level 1 
began to be deposited, at which time the rate of trash disposal increased 
dramatically. This sequence of events is better explained if Feature F29 is 
considered as originally being dug for use as a clay extraction pit, rather 
than purely a trash pit. 

The area to the north and northeast of the excavated block was mechanically 
stripped. Structure 75C was located in this area and was associated with 
Feature 75BF33, which may have represented the remains of another structure. 
Structure 75C was one of the few square or nearly square structures found at 
any of the sites. Its small size, posthole construction, and shape may indi- 
cate it had a function different from the majority of other structures. This 
function is hypothesized to be a shed rather than a domestic structure. Un- 
fortunately, time did not allow hand excavation of the structure or excava- 
tion of its postholes. Artifacts from the large feature to the south, F33, 
may reflect activities at Structure 75C, but this is only conjecture. 



124 




in Ci 

o :*: u_ 

CO 0) 

LU Ol i_ 

3 0> +J 

»-* -i- CJ 

Li_ CO U_ 



1 < 


1 03 


i: ^ 


i- oj 


4> 


4> 






o « 


O » 


*5 


^S 


c- 1 


_l 


5 "" " 


J 


O "O 


21 


^ c 


03 a 


03 O 


u> 


</> 






%3 




a2 


'3 2 



LU 



!T*T 



o 








or 






4> 


a. 


c 
a 


c 
o 

0) 


Sand 

( Lev 


< 

LU 


tn 


2 


o o 




c 


o 


is 




J 


s 




o 


• 


?a 




w 


>• 


55 




a) 








^m 






JC 


j= 


■*£ 




a 
Q 


a» 
'J 


a x 



&£ 



125 



Feature F33 was a layer of dark organic soil resting on the red clay subsoil. 
As seen from the trench excavated through it (Figure 41), the feature fea- 
thered out on its edges. After analysis in the laboratory and examination of 
comparative data from Feature F5 (the floor scatter covering Structure 75B2), 
it became apparent that F33 represented a floor scatter. The low number of 
artifacts (Table 12), their generally small and eroded nature, and the extent 
and shallow depth of much of F33 all pointed in this direction. 

Examination of Figure 41 shows that there was also a low depression in the 
subsoil near the eastern end of the trench. This may have represented a clay 
extraction/trash feature, which was subsequently covered by the floor 
scatter. The greatest depth of the depression into red clay subsoil was 2.0 
feet, as deep as any of the clay extraction/trash features near Structures 
75Bx and 75B2* In this depression were two decomposed posts, only one of 
which was excavated. Analysis has shown the wood to be of the pine family, 
but species identification was impossible due to the condition of the ma- 
terial. The distance between the two posts was 7.75 to 8 feet, much greater 
than the post distances at Structure 75C to the northeast. The only similar 
distance between posts at Feature F33 and at Structure 75C was achieved by 
measuring posts on adjacent walls. Indeed, the two posts in Feature F33 may 
have represented posts on adjacent walls of a second structure. 

Features F30, F31, and F32 made up a complex of three clay extraction pits, 
which were located 50 feet to the north of Structure 75C and over 100 feet to 
the northeast of Structures 75B^ and 75B2* These features were found during 
mechanical stripping of the northern portion of Site 38BK75 within the proj- 
ect right-of-way. The soil was stripped to within .1 or .2 foot above red 
clay subsoil by a bottom loading pan and then shovel shaved to define the 
features. The western half of each feature was excavated, a profile drawn, 
and finally the eastern halves were excavated. 

Features F30 and F31 extended 1.1 feet into subsoil, and F32 extended 1.6 
feet into the red clay (Figure 42). The overburden ranged from approximately 
.6 to 1.0 foot deep in this area so that the features were originally dug 
approximately 1.7 to 2.6 feet below surface. The overlapping of the features 
was obviously greater above the red clay layer than below it. This would not 
tend to happen if the features had been intentionally dug for trash pits, 
since digging up garbage to bury garbage would be highly unusual behavior. 
The fact that the features only overlapped to any great extent above the clay 
layer indicates that the objective of the pits was the clay layer itself. 

The artifacts retrieved from Features F31 and F32 showed similar Kitchen 
Group percentages and those from F30 and F32 had similar Architecture Group 
percentages (Table 11). Given that only 342 artifacts were retrieved from 
the features, any differences in pattern can be accounted for by small sample 
size. Rough calculations of volume were developed to determine arti- 
facts/cubic foot values. Feature F30 had 1.6/cubic foot, F31 had 3.2/cubic 
foot, and F31, the deepest pit, had 0.6/cubic foot. These low figures, the 
presence of water laid sand in the bottom of F31, and the intrusion into the 
clay layer indicated that the pits were not primarily meant for trash dis- 
posal and had been left open over a period of time and filled primarily by 
natural causes. 



126 



SSw 



II 










in 


ro 




r~- 


on 


1 — 1 


^ 


U_ 


^r 


CQ 






cc 


0) 


UJ 


m 


s_ 


C£ 




3 


3 


O) 


4-J 


C5 


-t-i 


ra 


1-1 


•^ 


<U 



00 



[ 



127 




CM 
fO 








CM 








oo 








U_ 








"o 








c 








<T3 








,_ 








ro 


O) 






Lu 


^ 






•< 


4- 






o 


o 




un 


on 


s_ 




r-- 


u_ 


c_ 


C\J 


^ 






«3" 


CD 


in 


^3 




CO 


0) 


C 


LU 


on 


'— 


rO 


Ci 




13 




3 


CD 


+-> 


C 


cs 


-t-> 


<T3 


ro 


^-* 


•r— 


CD 


f^ 


Lu 


tn 


U_ 


Q_ 



m o- 



§? £ ° > 

o- *</, »| 
•* It, 2 § 



i % 



u. 



■c - * ■ - 



ji 



CM 
(O 

u. 

I 

m 



o 
z 
< 



to 

Ik 

i 

m 

O 
to 

ik 






■o o 








c -C 








O CJ 






u. 


</J 


£ _ 




_l 




^5 




< 
I 


w C 


• 






o o 


■9 J 






■o 2 


(J 


5 

3 


z 
<r 

LU 


* 


» JJ 


</l 


1- 


g u 


o u. 

*• — 


o 


LU 


£ 


> o 


U 


$ 




^ (J 






i 5 

° i 
a x 


-J u 


• 


o 



o 

m 

I 

m 



o 

Q. 



t- 

< 
LU 



128 



The structure in the dirt road between 38BK75 and 38BK76 (Figure 26) was 
found during grading of an access road for the project. Since this area was 
outside the archaeological project area agreed upon prior to mitigation, the 
area was not mechanically stripped, as was the majority of Site 38BK75. Road 
grading only exposed one trench, which was not excavated, and this feature, 
with associated postholes, was covered with red clay road fill within an hour 
of being exposed. A hypothetical reconstruction of the second trench noted 
for this structure, based on extensive experience from Sites 38BK76 and 
388K245, is presented in Figure 26. Since this was a reconstruction, no 
dimensional data beyond length could be given. The structure had an align- 
ment approximately 90° off the other structures at 38BK75 and was similar to 
that of Structure 76L, which was oriented approximately 90° from the other 
structures at 38BK76. Whether this implies a similar function for the two 
structures or was simply personal taste on the part of the builders is im- 
possible to say, since both structures were discovered during mechanical 
stripping and the house in the road at 38BK75 had only one wall exposed. 

Mention should be made of a concentration of postholes to the northeast and 
just outside of the excavated block of Site 38BK75. This concentration was 
found during mechanical stripping and then mapped after shovel shaving. One 
apparently large feature was actually a place where the red clay subsoil 
naturally protruded an inch or so into the overlying sand layer and was, 
therefore, not a cultural feature. The posthole concentration itself had 
apparent regularities, but it proved impossible to distinguish a clear three 
or four sided structure in the concentration. Similarities of posthole 
shape, size, and fill were used in the field in an attempt to outline one or 
more structures, but without success. None of the features could be excavat- 
ed during mitigation, but it is unlikely that comparisons of arti factual or 
chemical soil data from the postholes would be of much help in establishing a 
structure. This area may have represented a two-sided structure or a special 
use area, or perhaps animal pens or posts associated with another plantation 
activity. 

Other postholes at the site appeared to be linear and were actually postmolds 
rather than holes. These undoubtedly represented fence lines as did the two 
postmolds in Unit 16 of the excavation block discussed above; however, it was 
impossible to date such features even though it is doubtful they were associ- 
ated with the structures. 

To the south of the excavated blocks were four trash features which were 
found at the end of work on 38BK75 and were therefore left unexcavated 
(Figure 29). With these features, Structures 75B^ and 75B? were surrounded 
on all sides by such features. This is not a pattern followed at either 
38BK76, 38BK245, or apparently at Spiers Landing (Drucker and Anthony 
1979:91). 

It is worth mentioning that South' s (1977:179-182) assumption that "odori me- 
tric" scaling is a deciding factor in discard behavior (also embraced by Le- 
wis 1981) was not borne out at Yaughan or Curriboo. Feature F29, the feature 
with the most bone, was placed near the structures while features further 
away had little or no bone and few artifacts. This holds true with some 
variation at Site 38BK76 and 38BK245, as well. Drucker and Anthony (1979) 
note the same variation from South' s scale at Spiers Landing. 



129 



Summary of Site 38BK75 

Briefly summarized, excavation of this site produced five certain structures 
and two possible structures (the posthole concentration just mentioned and 
the possible floor scatter at F33). All but two of the five structures were 
posthole structures, and the two possible structures were also of posthole 
construction. The fill of the trench house at 75B2 was sandy topsoil and 
unlike the trench fill elsewhere at Yaughan and Curriboo, but similar to post- 
hole fill. A floor scatter covered Structure 75B2 and indicated that a di- 
viding wall possibly ran down the center of the structure. A hearth outside 
of Structure 758^ was the only definite hearth found at 38BK75, 388K76, and 
the slave quarters at 38BK245. The remaining non-structural features were 
fence posts and clay extraction pits subsequently filled with low amounts of 
trash. The fence posts may or may not have corresponded to the slave occu- 
pation of the site. 

Site 38BK76 

Site 38BK76 was located several hundred feet to the northeast of Site 38BK75. 
The site was lower than 38BK75 and more poorly drained; the western end of 
38BK76 was inundated in the spring of 1979. The structures at 38BK76 were 
laid out along a slight and ill-defined rise which began at the road separ- 
ating the two sites and formed a "U" terminating at Structure 76B (Figure 
25). The following discussion begins at the westernmost structure, 76C, and 
continues across the site to Structure 76B. 

The structures and features at Site 38BK76 (Figure 26) are described in tabu- 
lar form on Tables 13 and 14. Artifacts are summarized on Table 15. The 
general stratigraphy of the site will be illustrated in detail in the discus- 
sions of Structures 76A and 76B where controlled block excavations were con- 
ducted. Structure 76C was aligned with the complex of structures termed 76DM 
to the southeast. This structure was not excavated and not much can be 
stated about the structure beyond what is presented in Figure 43. 

Feature Fl, located to the northwest of Structure 76C, was nearly rectangu- 
lar, filled with dark brown sand and extended .64 foot into subsoil. It 
contained few artifacts (Table 15). Because of the low artifact count and 
the rather shallow depth of the feature, it did not appear to have been 
primarily either a clay extraction or trash pit. Since this feature was 
found after mechanical stripping, it 1s possible that clues to its function 
may have been destroyed. 

Feature F2 was located adjacent to Feature Fll, to the south of Structure 76C 
and west of Structure 76DM. It extended 1.2 feet into subsoil . The artifact 
to volume ratio was 16.96/cubic foot, which was higher than many of the fea- 
tures at 38BK75 and high in comparison to most of the features at 38BK76. 
The majority of the non-Colono Kitchen Group artifacts were olive green glass 
fragments, which provided a total of 22 sherds more than the Clothing, Perso- 
nal, and Activities Groups combined, and accounted for more than half the 
Kitchen Group artifacts after Colono was removed. The only other feature 
with such a high proportion of bottle glass was Feature F8 near Structure 
76E, which had three restorable bottles. The nearly neutral pH of the soil 



130 



CL O 

*-> o 



—I oo —I 



O 

■43 






—I LP 



C\J — I 



— I — I LT> 



o 
o 



00 

o 



U ai 



■— c_> oi 



(II 



o 


. — — « 


■ — ' 


o 




Ol 




■o p*> 






O) 


u 


4-J 


t. — 


^"O 


■*-» 


,— 


c 


»o 


=i 


1- o 


*3 


o 


HJ 


*J 


*j •— 


at -c 


4-> 


ar 


4-> 


c 


u — 


» *j 


e 




^ 


Ol 


3 -r- 


O 01 


0) 


w 


•^ 


F— 


i- U. 


u S 


■^ 


i/j 


o 


E. 


*J 


0) 


l_ 


o 








go > 
» (J 

,— X 

■— 0) 

o 

in -O 

a. c 
o <o 



■o c 

«3 > 

(J « 

■>- >, O 

e « x 

TJ — UJ 

V. u 
Ol 

UT3 Ol 

Ol c 

C 1- — 

0) -O CL 

•— 01 T3 Q. 

o e oi ■«- 

4- -T- 4-J 

4J O E <« 

in 

(/i n »— 

ii <^ -.— 

43 - c 

a. >> « 

v u r 

- o — o 

u E 

C 4-> -O 

01 X Ol 11 
L tl 1. 

+-> *J Q. 

II ••- 

II 01 '- 

Ol tJ * J 



— < Co m t 



131 



a 

33 



o 






33 



i 



c c 
o o 



r-~ \n 



o 
a 



c 

J3 



Cl 

o 



a. c 
o <o 



o 

d 



o 
oo 



T3 II 

c 
« > 



o 

CO 



-o en 

01 c 






— i — o 



♦J LT) 



■u O S m 



o */> 


II 


| 


a.— 




T3 


VI 


s: 


U 


ii n 






.a 


■ 


c 


a. 


>< 


T3 


b. 


m 


JZ. 


- o 


1— 


u 


£. <*- 


'-) 


OJ 



••- ^ !"*> 



o u 

•— C_> 41 



4-> 


1- — 


^■O 


•C 


3 


!- O 


W 


4-* .— 


U £ 


C 


u — 


> 4-> 


4> 


3 ■— 


o ai 


•~» 


U U- 


o z: 


1. 


4-> 


ai 


O 


iy> 


ce 





01 


ai 


<-> 


, — 


c 


o 


-o 


X 


Jj 




VI 


*-> 




^1 


o 


o 








•o •— 



4i x 41 n 
U 41 U 

*j ■>-> a. 



— cm m t 



132 



















>i 




























T3 




































































>> 




>, 








>1 




<TJ 






>^ 




>. 


■c 




-o 








■o 




1/1 


c 




■a 




■a 


c 




c 








c 






I 




c 




e 


ia 




"0 








-a 




■a 


o .— 




T3 




'O 


vi 




v» 




~ 


£ 


vi 




0) 


1- — 
-a o 




VI 




v» 


e 




c 




o 


o 


c 




4-> 


V) 




C 




c 


5 




s 




vi 


VI 


3 




*J 


j* a. 




5 




I 


o 




o 




a. 


a. 


o 




o 


s- o 




o 




o 


■_ 


pa 


i. 


^_ 


o 


o 


L. 


r— 


e ^- 


TJ *J 


1— 


(. 


,— 


•- — 


-O 


•1— 


.13 


-r- 


«-> 


4-J 


n 


'fm 


■^ 


T 


■^ 


JZ 


•*" 


.a •— 




o 




O 








o 


c o 


>> 


o 




o 


o 


M 


v> 


_» 


V> 


>> 


>> 


-JC 


V) 


S VI 


>>-o 


VI 


-* 


V) 


•>• Vi 


s- 


a. 


s. 


a. 


01 


■V 


1_ 


CL 


o a. 


s- c 


a. 


t. 


a. 


-_ a. 



ho vi u 

= >*- O a 

U — o 



■o o 

ai vi 

.c a: J3 

*-> 3 

a. o <-n 



VI •»- 



«0 

a 






O 

a. 






133 



at ** — 

<n T3 3 •«-> 

= *- o o 

- — o 

*J ■ w i- U.I 

v> U u 



ai 





c 






in 


i/i 


c 


T3 






CL 




2 


ai 




T3 


O 


>> 


O r— 


L. 




e 


4-> 




S- ■— 


O 




tj 




"5 


-Q O 






</> 


>> 


T3 


i/> 


C 






-o 




J* a. 


2 




>l 


e 


>s 


<- O 


o 




f 


"3 


01 


TJ •!-> 


L. 




1/1 


■XI 


L. 


■a 


a 


• ,_ 


"O 




CTl 


>, 




o 




C 




>*-o 


.*. 


in 


>, 


2 


_* 


s- c 


4. 


a. 


01 


o 


L. 


1) T> 


<n 


o 


s_ 


L. 


-o 


^ 1/1 


■a 


4-> 


C7> 


-Q 


T3 



■o 


o 




01 


•> 




-c a: 


.a 




w 


3 


o 


Q. o 


iy> 


^5 


V j-> 






o c 


^ 




— ' 


'O 





2 
>s O 



0) 

c 



CM 
•0 





^O 




VO 


CM 


r- 


<*■ 


p^ 


X. 


01 


t/\ 


01 




u 


Ol 


L. 


•V 


3 


»— 


3 


L. 


4-> 


o 


4-1 


3 


<J 


£ 


<J 


u 


3 


*J 


3 


•W 


i. 


tn 


1. 


0) 


*J 


o 


«J 


LL. 


t/i 


a. 


GO 



— ' CM 



I 



134 



_ _ Q 



F 



— o 

o 



>3- O — i 





o 


o 




o 


o 


** 








o 


o 


_* 


in 


u-> 


jj. 






,^o 






h 







>, 




.Q 






Il- 


<y* 


ia 


c 

L. 


£r 




h- 



w 



I? 



IS 



_ — i o 



—i — o 



CD 



CM 

ao 



•^ r» cm 



CM — • 



■ — I 



S 
<: 



— CO 



CO — 
^» ..li ii ii 

•& <»> 

— . — iOOOOOO 



o o 

o o 



* 



o 

2 I 



^ o o o o o 



in 



a? — 
as — 



LTltMOOOOOO 






OCVJOOOO— • o 



CO 



g 



— <00000<MO 



o — 



o 
o 



O — i — I 



is 



a. 
o 



i« ._ 



5. I 



135 



p? 



is 



F 



— r- 



F 



15 



00 CM 

U1 — . 



O 

© 



— ooooo©© 



o 
(J 



II 



•■— l/> 



o 
o 






: 



136 



ft 



137 



o 



00 

a O 



a k 



\o 






G o 0. 











i/> 








'_ 


0) 








VO 


s_ 








r^~ 


3 






l/l 




+J 






O) 


-o 


ro 






s. 


c 


QJ 






3 


(O 


Li_ 






+J 










u 


#1 


"O 






3 


XI 


0) 






s_ 


vr> 


•*-> 




vO 


■H 


r-. 


03 




r-~ 


on 




°^- 


ro 


^ 




r. 


U 


^r 


C2 


<4- 


Q 


o 




CC 


o 


v£> 


00 


UJ 


m 




l^» 


W) 


Q£ 




I/) 




«C 


— 


(U 


c 


#» 




O 


-4-> 


(0 


O 


-a 


i— i 


•r» 


r— 


US 


c 


U_ 


</l 


D- 


r-» 


ra 



<5> 



Or 



§ 





o 











o 

o 



jE 




o. 



138 



indicates that bone could have been preserved in the feature, although no 
bone was recovered. Considering its depth and its relatively high amount of 
artifacts, this feature may have represented a clay extraction pit, which was 
later intensively used as a trash pit. 

Feature Fll was very shallow, extending only .28 foot into subsoil. Artifact 
density could not be determined with any accuracy. Its location next to F2 
may indicate it was dug and filled with debris after F2 was full. Its shal- 
low depth indicated that it probably functioned primarily as a trash pit. 

Structure 76M was located to the southeast of Structure 76C (Figure 43). 
Outside the southwest wall of Structure 76M was a row of postholes parallel 
to the wall. This row of posts was approximately half the distance from the 
southwest wall than the southwest wall was to the northeast wall. This 
spacing and the fact that no other structures were as narrow as the distance 
between the extra row of postholes and the southwest wall have led to the 
conclusion that this row represented a porch or extra room. Fill in all the 
postholes of Structure 76M was essentially the same, topsoil mixed with clay 
subsoil, which would indicate that all postholes at the structure were ex- 
cavated at the same time. Structure 76M was aligned with Structure 76D, 
although separated from it on the southeast end by five or six feet. It is 
possible that the two structures were attached by a breezeway, although this 
is unlikely, since there are no extra support posts, which would be expected 
if a breezeway were present. 

The artifacts summarized in Table 15 constitute all artifacts from the area 
around Structures 76C, 76D, and 76M, including surface collections, exca- 
vation of numbered features noted on Figure 43, and excavation of the 
southern trench at 76D. 

Structure 76D consisted of a trench house (76D^) to which a larger posthole 
addition (76D2) was later added (Figure 43). This was made clear in the 
field by the nearly identical widths and alignments of the trench and post- 
hole foundations, and by the replacement on the northeast trench of two posts 
by postholes with fill identical to that of the posthole structure. Apparent- 
ly, when the posthole addition was added to Structure 76D^ , certain posts of 
the original trench house were repaired. 

The posthole addition also appeared to have undergone repair on the northeast 
wall as shown by an extra posthole just inside of the second posthole from 
the north corner of the structure. This was one of the very few examples of 
extra, possibly replacement posts being placed inside a structure. The ad- 
joining wall between the posthole addition and the trench house may have been 
repaired also as witnessed by two extra postholes around the central post- 
hole, one to the inside, and the other along the wall. There was also an 
extra posthole next to the last posthole mentioned and just inside the trench 
house indicating that this portion of the structure may have received extra 
support. 

The opening between the trenchs at the southeastern end of the Structure 76D 
trench house had a posthole which was off-center. This may have represented 
a door post as well as structural support. Often off-center postholes were 
found at both ends of trench structures, perhaps implying two entrances. 



139 



At Structure 76D, there were also central postholes down the center of the 
trench structure. Three of these were well aligned and perhaps represented 
floor supports. The fourth may have represented repair or added support. 
Examination of the other structures at 38BK76 shows that no other structures 
at Yaughan had such postholes. This indicates that either the other houses 
had wood floors, but they were not supported in the center, or that they had 
earth floors. Evidence given below suggests that they probably had earth 
floors. 

The posthole half of Structure 76D has no central postholes. Since it can be 
shown that both halves of Structure 76D were used simultaneously and the 
interior construction differed between halves, there may have been functional 
differences as well. The trench half may have been used for habitation and 
the posthole addition for storage or some other function. Conclusions on the 
overall function of the Structure 76D complex will be discussed in detail 
later in this chapter. 

Associated with Structures 76D and 76M were four large features, F3, F4, F10, 
and F5. The first three were associated more closely with Structure 76M and 
the last with Structure 76D. Feature F3 was square, shallow, and filled with 
sandy gray organic soil. Feature F4 was virtually identical in depth, shape, 
and fill, although smaller. Feature F10 was larger, but it was also square, 
shallow, and filled with sandy gray organic soil. 

These features formed a line outside and parallel to the southwest wall of 
Structure 76M. Features F4, F10, and to a less clear extent, F3 were also 
oriented in the same manner as Structure 76M. The nearly square shape of F4 
and F10 and their locations near the corners of the structure made them 
different from any other features at 38BK76. During excavation, it had been 
hoped that they represented wells or privies. Feature F4 extended .3 foot 
into subsoil, F10 extended .65 foot, and F3 only .25 foot, thereby 
eliminating that likelihood. 

Chemical soil analysis was run on Feature F10, as it was the most clearly 
defined and deepest feature (Appendix C). It exhibited significantly less 
nitrogen and more phosphorus than the natural soil matrix range. Coupled 
with a nearly neutral pH, the potential for the presence and preservation of 
bone was good, but no bone was found in any of the three features. The vir- 
tual absence of artifacts and lack of bone where preservation should have 
been relatively good indicated that these features were not primarily or 
perhaps even secondarily used as trash pits. The shallowness of the features 
ruled out clay extraction pits, privies, and wells, and the absence of post- 
molds ruled out large postholes. The features may have been excavated for 
small gardens or flower beds. 

Feature F5 was also found during mechanical stripping and may have been 
larger horizontally than is recorded here, since it appears to have feathered 
out on its northern edge. This feathering may have been due to a logging 
road which ran along the northern edge of the feature and caused it to be 
spread in that direction. At its deepest point, the feature extended .71 
foot into red clay subsoil. The irregular nature of the feature made even 
rough computation of artifacts per cubic foot impossible. Chemical soil 
analysis indicated that all values were within the range of the normal soil 



140 



matrix (Appendix C). No bone was recovered, and the function of the feature 
is unclear. 

Feature F7, located midway between Structures 76DM and 76E to the northeast, 
extended .43 foot into subsoil. Only the western half of this feature was 
excavated and profiled. This feature may also have been impacted on the 
surface by a logging road. Chemical soil analysis showed a normal profile 
(Appendix C) and no bone was recovered. A rough calculation yielded approx- 
imately 20 artifacts per cubic foot, which was higher than that for Feature 
F2 near Structures 76C and 76DM. The shallowness of the feature and its 
relatively high density of artifacts, especially Kitchen Group artifacts, 
indicated that the pit may have been used primarily as a trash feature. 

The location of this feature almost exactly midway between 76DM and 76E made 
it impossible to show association with either of the structures over the 
other. The area immediately around the feature had a cluster of postholes in 
which no structural alignments could be determined (Figure 43). One reason 
for this may have been that the area was impacted by logging vehicles in the 
recent past. Although logging roads occurred over the site, this area ap- 
peared to be randomly impacted by heavy equipment or tractor ruts which may 
have obliterated key postholes for a structure prior to mitigation. 

Feature F15 was observed after a brick was dislodged by the motor grader. 
Mechanical stripping near the feature ceased at this point, and the brick 
fragment was replaced. It was hoped that the brick would represent a pier 
support; however, this was not the case. The feature was a shallow (.45 foot 
into subsoil) square pit with large brick fragments imbedded in the surface. 
This was the only feature with large brick fragments at Site 38BK76. Un- 
fortunately, no width to length to height ratios could be developed for the 
bricks because of their fragmentary nature, and the function of the feature 
could not be determined. 

Structure 76E (Figure 43), located northeast of the structures and features 
discussed previously, had a central trench perpendicular to the axis of the 
structure which divided it into halves. There was also a partial trench on 
the southeast end of the structure. The open space remaining at the end of 
this end trench was less than half the width of the structure and seemed to 
correspond to the off-center postholes on the ends of other trench struc- 
tures, such as, Structure 76D discussed above. This indicated that the open 
space on Structure 76E probably represented a doorway and that the shorter 
spaces on structures with off-center postholes also probably represented 
doorways. 

This was the only structure at 38BK76 which had central and end trenches. It 
could not be determined if the structure also had a trench on the northwest 
end, since this end of the structure had been heavily impacted by logging 
equipment during a period when the area was saturated by groundwater. It 
should be noted that the perpendicular trenches in this structure did not 
touch as they did on the double bay structures at 38BK245, nor were they as 
regular in outline, width, or alignment as the structures at 388K245. 



141 



There was a parallel line of three postholes across the width of the struc- 
ture, in addition to the central wall. These were not placed in an optimum 
location for floor support, as they were too close to the central wall and 
probably served a different purpose. Similar lines of postholes were found 
at the barn and two double bay structures at Site 38BK245. These lines of 
postholes may have represented supports for interior walls, shelves, 
"built-in" tables, beds, or for hanging utensils. At the hypothesized barn 
at Site 38BK245, the line of postholes on the northeastern end of the struc- 
ture were thought to have represented a wall separating a storage room from 
the larger area and perhaps served as loft supports, as well. 

Feature F8, associated with Structure 76E (Figure 43), was distinctively 
square at the surface, apparently deep, and filled with a very dark organic 
sand, unlike the other features at 38BK76. It was first thought to be a 
well, but was soon discovered to extend only 1.5 feet into subsoil. The 
sides were neatly and almost vertically dug and the bottom was flat. Three 
reconstructive olive green wine bottles were recovered from the feature. 
The MCD for this feature is 1773.6, based on ten datable sherds. 

The two main levels in Feature F8 were separated by a layer of water laid 
light yellow sand (Figure 44). Two of the restorable bottles came from Level 
1. These were a sand pontil English style bottle and a glass or rod pontil 
French style bottle (Jones 1971:68-69). The third bottle was made up of 
sherds from Levels 1 and 2 and was a sand pontil English style bottle. 

Although the pH of the feature fill was slightly acid, the total organic 
phosphate and carbon readings were significantly high (Appendix C). This was 
undoubtedly due to the relatively elevated amount of bone, 7.01 grams, found 
at this feature. Even though this was insignificant compared to Site 
38BK245, it was the most recorded from any single feature at Site 38BK76. 
The two levels may have represented slightly different filling episodes, but 
the cross-mended bottle and the similarity in chemistry and soil color indi- 
cate that the time difference was probably short and did not represent any 
change in function during the pit's refill period. 

The artifact to cubic feet of fill ratio of 51 artifacts per cubic foot for 
Feature F8 was more accurate than most, because of the feature's regularity 
of shape. This was the highest ratio of any feature at Site 38BK76 and 
overshadowed all features at the other sites. 

The placement of the feature within 6.5 feet of the hypothesized door of 76F, 
its regular shape, the care with which it was dug, its depth, its artifact 
density, and high chemical values indicate that the feature fulfilled a spe- 
cialized function, perhaps a privy of short duration. If this is true, it is 
contrary to South's (1977:179-182) "odorimetric" scaling hypothesis. 



142 



EAST PROFILE 



GRADED SURFACE 




Dark Brown Sand 
Level I 



Yellowish Brown Sandy 
Rain Wash Fill 



Light Yellow Sand Lens 



Red Clay Subsoil 






Very Dark Brown Sand 
Level 2 



Rock 




TREE ROOT 
DISTURBANCE 



SURFACE OUTLINE 



FIGURE 44 
Site 38BK75 
Feature F8 



143 



There were no artifacts recovered which were directly attributable to 
Structure 76F (Figure 45), the next structure to the east. This structure 
had four parallel trenches. The southernmost trenches were assumed to be the 
main structure, based on size and space of the area enclosed by them. The 
two northern trenches enclosed an area too small and narrow to be a separate 
structure and possibly represented a shed addition or perhaps replacement 
trenches. The northeastern trench of the main structure extended to the 
northwest, was narrower than the remainder of the trench and appeared to 
correspond to three postholes at that end of the structure, forming a square. 
This square may have represented a porch or an additional shed. No trash 
features were found in association with the structure. 

Structure 76G was located to the southeast of Structure 76F. It consisted of 
three trenches and two clusters of postholes. The main portion of the 
structure included the two northeastern trenches. The third trench was 
narrower and shorter than the other two and it is unclear whether this trench 
represented an additional room, a shed, or a porch. The postholes on the 
northwest end may have represented a porch, but those on the southeastern end 
are too random to form a pattern. 

The feature to the southeast of Structure 76G (Figure 45) was a shallow 
discoloration of the soil which may have represented the very bottom of a 
shallow feature. This possible feature was not excavated when its depth 
(less than one inch) was determined. 

Structure 761 appears to be a three-sided, square, posthole structure located 
northeast of and apparently aligned with Structure 76G. There were three 
additional postholes clustering around the east corner of Structure 761, but 
their function did not appear to be structural (Figure 45). No artifacts 
directly attributable to Structure 761 were recovered. It was concluded that 
this structure, which lacked a northwest wall, functioned as a shed or 
special use structure related to plantation activities. 

Feature F24 was shallow and thought at first to be a trash feature associated 
with Structure 761, but no artifacts were found. Similar to the "features" 
in the cluster of features southwest of Structure 76G, its function is 
unknown. 

Structure 76J, east of Structures 76G and 761 and north of 76A, had three 
posts along the southeast side, three on the northeast side, and three on the 
southwest side (counting corner posts twice). The northwest side appeared to 
be open, similar to Structure 761. Other postholes in and around the 
structure made identification difficult, as fill in all the postholes was of 
the same color and texture, indicating contemporaneity. 

It is conceivable that the two postholes outside the structure to the east 
and west supported a ridge pole for a roof or hanging utensils. If this is 
true, it argues for an open-sided and covered work area, since an enclosed 
structure would presumably have the roof supported from the inside. In any 
case, no artifacts 1n association with Structure 76J were found, nor were any 
features suggesting a special function such as blacksmi thing, cooking, or 
indigo production. 



144 



2 

O 
O 

CO 




UJ 




O <b 



<3> 
















O 



o 






0°. 



o r^ 



-a 

c 

13 
00 " 

■M 

U " 

3 1-1 

S- <J=> 

iO +J is 
rs oo 

CO o« 
uj on r»» 
ca oo 

3 0J C * 

CJ -M .-o — 

U. 1/) O.N 



0J 



u 

o 

oo 

< 

— 
C 

'■a 




o 

o 

CM 



o 








o 



0«£ 



fO 



o 



c<> 



-U7 U- 



o 



<? 



<© 

f- 








/> 



CD 



145 



Structure 76H was designated a structure for the purposes of analysis, since 
it presented a cluster of large rectangular features. Only Features F12 and 
F13 had any depth, however, and they were much larger than any structural 
postholes. The function of this series of features is unknown. 

Feature F12 (Figure 46) extended .8 foot into subsoil. Chemical soil analy- 
sis showed that carbon was significantly high and total organic nitrogen and 
total organic phosphate were within the normal soil range (Appendix C). This 
was due in large part to the remains of a charred basket found on the bottom 
of the feature. The basket was first identified during excavation of the 
west half of the pit in the section profile. The hypothesized area of the 
basket, approximately one foot by one foot, was pedestaled and a cookie sheet 
was inserted under the pedestal. The entire pedestal and cookie sheet were 
then slid onto a 3/4-inch plywood board, packed in aluminum foil, and re- 
turned to the Marietta laboratory. During the analysis phase, the feature 
was carefully hand excavated in the laboratory. The part of the basket 
exposed in the cross-section did not extend more than 1/4-inch into the 
pedestal. Since the basket had appeared to be collapsed in profile, it had 
been hoped that it would extend further into the pedestal, allowing for 
comparison with modern Gull ah basketry. This proved impossible, however. 
From field examination, the basket appeared to have been grass bundles sewn 
together in coils, as are Gullah baskets. Since the in- situ position of the 
basket was such that close-up photographs were impossible, photographs were 
taken from two to three feet away; these were not detailed enough to provide 
additional information. 

Feature F13 extended .6 foot into red clay subsoil and was semicircular in 
cross-section (Figure 47). Chemical soil analysis indicated that carbon was 
also slightly higher than the normal range for the natural soil at this site, 
although no organic material suggesting a basket was found. The low amount 
of artifacts made a case against the feature being a trash pit, but the high 
carbon and shallowness of the feature argue for such a determination, es- 
pecially since Feature F12 also had high carbon and evidence of trash dis- 
posal. The functions of the features were tentatively identified as a trash 
disposal, based on their fill and depth. 

Structure 76A, southeast of the previous structures and features (Figures 46, 
47, and 31) was located during excavation of Test Unit 3, when two postholes 
were encountered on what became the southeast wall. Excavation during miti- 
gation resulted in the first trench structure found on the project. Feature 
F40 was sectioned in order to examine its depth, form, and fill sequence. 
Fill was red clay with small amounts of topsoil and there was an indication 
of posts in the form of a darker vertical stain in the center of the feature. 
The excavation technique was changed from vertical sectioning to carefully 
excavating the trenches two or three inches below subsoil until the posthole 
pattern emerged. Postmolds were then excavated separately and mapped. 

Structure 76A was perhaps the most repaired structure at 38BK76. All corner 
posts on the trenches were replaced or repaired at least once (Figure 48). 
The function of the small perpendicular trench on the northwest end of the 



GRID 
NORTH 




146 



| 

SCALE IN FEET 



B 



m£ AUGER HOLE' '^ffM^M^ 



foy* AUGER mole : ^:x:::x:x:x>:>x:j 0, ^^::::x-::>x:x:x:::::x::: 

i ' l ' /X ' " 1 * ' ! 1 * 8 *!* !*!* ' ! ' !* ' ■' *- ■ * ' * ' l ''■■!• ' • ' ■ ' ■ ' I < ^ ' I * ! ' . ' ■'■• *•* ' ■' ' •* ' . ' .*.) I ''■''■ ' llv'. ; , i ,',K l i l;. ll .''|',' ' |t).J, 1 , 1 



A 



^^^~*t*-m 



Dark Organic Sandy Loam 
ore Intensely Organic Stained Loam 



Grayish Sandy Loam 



GRAOED 
SURFACE 



B^ 



' "'Wffv.vA' i' .v.v.v.'T'.vi l :!-"!v"^^ ! vX-^v!-"v!xlx-:-x-x-x-x 



AUGER 
HOLE 



CARBONIZED BASKET 



FIGURE 46 
Site 38BK76 
Feature F12 



147 



PLAN VIEW 





nj-n 

1/2 I 

SCALE IN FEET 



I3^g|| Vary Dark Brown 
&>?mSy Sandy Topsoil 



EAST PROFILE 



AUGER HOLE 




FIGURE 47 
Site 38BK75 
Feature F13 



148 




co ^ 

«sr cc 
co 

UJ CO 



> 

U 
X 

UJ 

_<- 
u 
o 



U. (/I (/) C2 



o 

a. 



c 



149 



house was at first thought to be the remains of a chimney foundation. Com- 
plete excavation showed this to be in error, and it is now recognized that 
the trench was an end wall similar to that at Structure 76E. 

When calculating length to width ratios and floor space noted in Table 13, 
the trench structure alone was used, as was the case with other structures 
containing additional walls or porches. The porch areas were added when 
multiple structures, two or more bays, are discussed later in this chapter. 

All bone from excavated units at 76A came from Levels 1 and 2, the root mat 
and the dark sandy topsoil (Figure 49). This consisted of 15.49 grams of 
mammal bone in Level 1 (perhaps from recent hunting activities) and 2.2 grams 
in Level 2 of which 1.9 was mammal bone and the remainder land snail. One 
posthole contained an additional 3.81 grams of oyster shell. 

The only possible hearth or trash feature at Structure 76A was Feature F33, 
which was shallow (.25 foot into subsoil) and filled with ashy soil (Figure 
50). The feature was deeper at the northwest end and feathered out towards 
the southeast. The amount of total organic carbon in the feature was more 
than twice two standard deviations above the mean for the surrounding soil 
(Appendix C). Carbon was also higher at this feature than at Feature F25 at 
38BK75, the hearth at Structures 75Bi and 75B2- No bone was recovered from 
the feature. There was no fired clay or large pieces of charcoal in the 
fill, which indicated that the feature may have been an ash dump rather than 
a hearth. The low amount of kitchen and colono artifacts (Table 15) also 
argues against the feature's use as an open air hearth. It is possible that 
the feature represented a soap making feature such as that found by Drucker 
and Anthony (1979) at Spiers Landing. This might explain the low amount of 
Kitchen Group artifacts. 

Structure 76K, south of Structure 76A, was a typical trench structure with 
postholes near one end, but there were two parallel trenches on the south- 
western wall (Figure 51). Where these two trenches touched or nearly 
touched, it appeared that one of the trenches was built as a replacement; 
however, a logging road had heavily damaged the trenches, making interpreta- 
tion impossible. The house measurements were taken from the outside trenches 
which were more nearly parallel and of the the same length, and which were 
assumed to represent the original structure. To the northwest of the struc- 
ture were three very shallow stains which were not excavated, but may be 
similar in function to shallow features such as Feature F24 discussed above. 

Structure 76L, close to and southeast of Structure 76K, had postholes clus- 
tered on its northwest side and a smaller trench parallel to the main 
trenches. This trench may have represented a shed addition to the structure, 
since it was much shorter than the other trenches and far enough away to 
allow for usable space between it and the main structure. The two main 
trenches were used for house measurements. This structure was oriented at 
90° to the other structures at 38BK76, which may imply a different function, 
or simply personal taste of the builder, if such was allowed by the owner. 
Orientation of all structures is discussed in detail at the end of this 
chapter. 



150 



i: 



lo 

r*s 

en ^ 

t ca 

CO 

lu en 
=> CJ 

O 4-J 



cj 

U 



lJ- OO 00 CQ 



o 

SL 



5 •■ *J 



CM 

aO 

o" 

c\j 
aj 

en 

Z 

LU 

o 

< 
03 



a I 



1 



m 

CM 

cO 

c\J 

N-" 

<n 

t Si 

1 t 

z 
a. = 
O 

m 

>c 
_i 

< 

ca 



«o 



» I 




ce 

o 



<n 

< 

UJ 



! 5 I 

CD Q j 



z 

3 



151 



ao 
u_ 

UJ 

<r 

3 




UJ 
O 

a: 
a. 



O 

z 








CM 






CO 






Lu 






■a 






c 






w 






m 




U3 


CI 




r-~ 


u. 


o 


^ 




LO 


cc 


</i 




cc 


ai 


bJ 


c-o 


s_ 


OS 




3 


=3 


ai 


4-> 


CC 


+-> 


H3 


M 


•i— 


a> 



00 





c 



O 

w 



e 
o 

» 







I- 
< 
UJ 

u. 




UJ 

_l 

u. 

o 

cc 
a. 

»- 

< 

UJ 

X 
H- 

cc 
o 

z 



152 









10 






r>i 






"O 






c 






113 






1^ t/1 






<X> <l) 






r*. s_ 






2 






i/i — 






o; re 






s- a; 






3 Ll_ 






4-> 






u "a 






3 0) 






- — 




U3 


+J <ts 




l>s 


OO -r- 


^-t 


^ 


<J 


m 


ca 


4- O 




cc 


O <s> 


LU 


en 


to 


OH 




wi <c 


ZD 


CD 


c 


O 


■M 


(T3 "O 




U- OO Q. (T3 



a) 

O 

o 

+ 






O a 




(0 



o 



+ 



4 



153 



Feature F14, presumably associated with Structure 76L, yielded a nearly 
complete Colono pot. The pot was very friable and crudely made, and it 
resembled a beginning potter's attempt at a cooking pot. Most of the Kitchen 
Group artifacts noted in Table 15 belong to this pot. The feature extended a 
shallow .35 foot into subsoil. Chemical soil analysis showed significantly 
high amounts of carbon and phosphates (Appendix C), but no bone was 
recovered. The feature's high carbon reading, soil, shape, and the presence 
of a nearly complete pot were reminiscent of Feature F12 with its carbonized 
basket. Both were, therefore, identified as trash pits. 

Structure 76B, southwest of the two previous structures (Figure 52), was 
actually two structures superimposed on each other (Figure 53). This 
structure was found in the second and last hand excavated block at 38BK76 
(Figure 54). Stratigraphy at the structure is similar, but deeper than that 
at Structure 76A (Figure 55). Structure 76B2 was superimposed on Structure 
76Bi as indicated by intrusive postholes in the trenches of 76Bi • Although 
76Bi fit the pattern established for size and shape at all three sites, the 
larger and squarer Structure 76B2 did not fit the pattern and may have 
represented a separate barn, storage shed, or perhaps a very late 
manifestation of a slave cabin. There was a series of postholes along the 
southeastern walls of the structures which belonged to the trench structure 
(Figure 53). These postholes were outside a line perpendicular to the axis 
of the structures. Postholes F96 and F98 strikingly resembled the porches or 
stoops of the two structures at Structure 75B and those at Structure 76G. 
These postholes clearly aligned with Structure 76Bi, the trench structure, 
and indicated a more radical modification to the structure than the simple 
posthole patterns on the ends of many structures at 38BK76. Since there was 
no evidence of fired clay daub excavated in that part of the block, these 
posts probably represented a porch rather than a wattle and daub chimney. 

The mean ceramic date (MCD) for Structures 76Bi and 76B2* (1787.6) was later 
than that for Structure 76A (1773.4). Structure 76B was, in fact, closer to 
the MCD established for Structure 75B (1789.8). Since the overall MCD at 
Site 38BK76 was 1773.0, Structure 76B may have represented the latest 
occupation of the site. The much higher total of nonlocal ceramics at 
Structure 76B than at Structure 76A, 691 (7.0 percent) to 173 (2.8 percent), 
was also closer to the pattern set at Structure 75B (10.6 percent). This may 
have been due to the increasing availability of nonlocal ceramics over time 
at Yaughan plantation. This problem is discussed in more detail in the 
following chapters. 

Feature F82 (Figure 50) may have been a hearth. Chemical soil analysis of 
this feature showed that carbon was significantly high and nitrogen and 
phosphate were within the normal range (Appendix C). This was a replication 
of the data for the other possible hearth at Structure 76A, Feature F33. The 
artifact patterns were also similar. However, Feature F82 extended .80 foot 
into subsoil and was much larger than F33. Feature F82 was also located 
within the confines of Structure 76B and not outside the structure, as was 
Feature F33. Feature F82 seems to have been associated with the posthole 
structure, 76B2, rather than with the trench structure due to the feature's 



154 



CO 

O 

O 



CO 




o 


CO 


O 
CO 


U3 


1 


a 


4- 


s_ 


1 


3 




■»-> 




U 




3 




<sO S_ 




r-» +J 




CM M Ol 




LO CO 




CC <4- 




LU CO O 




eg 




3 u c 




C-P (C 



U_ OO Q_ 




CD 



+ 




155 



40W 



30W 



20 W 



I0W 



JON 




20 N 



ION 



© 



ru — l 



12 4 a 

SCALE IN FEET 



76 8, 
78 8. 



FIGURE 53 

Site 38BK76 Structure 

Block Excavation 



'1 



156 




STRUCTURE 76 8 
LOOKING NORTHWEST 




STRUCTURE 245 B 
LOOKING SOUTHWEST 



FIGURE 54 

Photos of Structure 76B 

and Structure 245B 



157 



S L 









*■>• 




C7> 

rO 

aO 

ao 

CM 



rO 
CO 

z 

3 



CO 

-I 
< 

CD 



CO 

< 
UJ 



z 












s*s 



CVJ 



CM 

t^" 

CM 

CO 

\- 

z 

3 

U. 
O 

CO 

-J 

< 
03 



3 
O 
CO 






\~-l 



turn! r. ■ 

SSSSL 



3 ttfS 



is: 



K 4 









e 










o 










•r~ 








CO 


+■> 








<J3 


<a 






UD 


1 — 


> 






r~~ 




03 




LP 


^ 


CD 


<_>• 




LO 


CO 


i. 


X 


l/l 




CO 


3 


LU 


OJ 


UJ 


m 


*-> 




1— 


a. 




U 


— 


• I— 


ZD 


■D 


3 


U 


4- 


(JZ 


4-> 


S_ 


o 


O 


t— i 


• ^ 


-t-> 


i— 


S. 


u_ 


CO 


oo 


CO 


Q_ 



3o 



* r l 

» I — a 





158 



location, which was more symmetrical in relation to Structure 76B2 than was 
the case with Structure 76Bi» Of the 31 Kitchen Group artifacts associated 
with the feature, 29 were colonoware and none were nonlocal ceramics. This 
tentatively argues for an earlier date for the feature and, hence, an asso- 
ciation with the earlier trench structure. Although Feature F33 at Structure 
76A had no seeds, Feature F82 yielded 73. Half of the seeds found in the 
feature were unidentifiable. Of those identified, two were rice, one was 
maize, five were legumes, and 37 were unidentified grasses. This feature may 
have been used as a hearth as the carbon and seeds suggest. However, no 
burned clay daub was found as would be expected in a hearth. It should be 
noted that this was only one of two features, besides postholes, found inside 
structures at 38BK76 or 38BK75. The other structure with an interior feature 
was Structure 76L, which contained a shallow round feature. This feature was 
not excavated, but was filled with brown organic topsoil , and not grey ashy 
fill. 

Summary of Site 38BK76 

This site produced 13 structures, three from hand excavated blocks and ten 
from mechanical stripping and shovel shaving. Eight structures were pri- 
marily or exclusively trench structures (76C, 76E, 76F, 76G, 76A, 76K, 76L, 
and 76Bi), one was half trench and half posthole (76D^ and 76D2), and four 
structures were exclusively of posthole construction (76M, 761, 76J, and 
76B2). It was impossible to determine whether the posthole concentration 
west of Structure 76E constituted a structure because of logging disturbance. 
Trench fill was primarily red clay with minor amounts of topsoil. Two pos- 
sible hearths were located, one outside of Structure 76A and one inside Struc- 
ture 76B. Neither possible hearth followed the pattern of posts at the 
hearth at Structure 75B, and both identifications were very tentative. The 
remaining features were clay extraction/trash pits, and possibly garden 
plots. Perhaps some of the shallower features were exclusively trash pits. 

It might be noted that clusters of three or four postholes appeared at the 
corners of some of the structures. These may have indicated poorly defined 
porches or stoops as opposed to the larger additions and/or porches at 
Structures 76M, 76A, 76B, 76G, and 76F. Also, Structures 76D and 76M were 
closely aligned and represented the single largest structural complex. 

Site 38BK245 

Site 38BK245 was located approximately three-quarters of a mile southeast of 
38BK75 and 38BK76 and was approximately the same distance from the Santee 
River (Figure 1). The layout of Site 38BK245 encompassed a large, flat, 
mechanically stripped area on the south, a large soil borrow pit in the 
center, and the original ground surface on the north (Figure 27). Most of 
the structures discussed below were located in the mechanically stripped 
area. These included 245A, 245B, 245D, 245E, 245F, 245G, and 245H. Struc- 
ture 245K was located on a pedestal left in the soil borrow pit and Structure 
245C was in the relatively untouched area north of the borrow pit. 



159 



The structures and features at 38BK245 (Figure 27) are described in tabular 
form on Tables 16 and 17. Artifacts are summarized on Table 18. Generally, 
the site had been heavily impacted prior to mitigation by mechanical strip- 
ping of the topsoil and freezing and thawing of the exposed subsoil. It is 
unknown how far such stripping extended into the subsoil, but it is clear 
that in some areas the impacts completely destroyed parts of features. This 
was particularly evident in the examination of the long trenches which often 
became shallower at the northeastern ends until they disappeared completely 
(Figure 27). The stratigraphy is assumed to have been similar to that still 
remaining at Structure 38BK245C, although the stratigraphy at that structure 
was undoubtedly affected by the great amounts of brick and mortar debris 
present there. Structure 38BK245C represented the only block excavation at 
388K245, although a cellar and brick clamp were hand excavated in quadrants. 
The remaining data came from mechanically stripped areas which were only 
cleaned up during mitigation and from the features exposed after shovel 
shaving. 

Since no dark brown or light yellow topsoil remained at the site, there was 
no standard with which to chemically compare the features in the stripped 
area at 388K245. The natural soil horizons at Structure 245C could not be 
used for this purpose since the area around Structure 245C was heavily 
impacted by brick and tabby mortar debris from the chimney and piers, which 
could have affected the soil chemistry. Therefore, comparisons were made 
between the features in the stripped area at 38BK245 and all other features 
in that area and significance levels at two standard deviations around the 
mean were established. 

The easternmost structure at the site was Structure 38BK245B. There was, 
however, a series of features to the northeast, of Structure 38BK245B, which 
may have represented another structure. The narrow trench in this location 
may have been either the remains of a drip line or an irrigation ditch, 
whereas the postholes suggested a structure similar to 38BK245G. 

Structure 38BK245B had two bays (Figure 56); Bay 1 was at the northeastern 
end of the structure and Bay 2 was at the southwestern end. The interior of 
Bay 1 exhibited two lines of postholes which may have been floor supports. 
The interior of Bay 2 had a line of three postholes along the southwest wall 
which may have reflected a function other than floor support. Replacement 
postholes were found along the exterior of the northwest and southeast walls 
and along the interior of the northwest wall in Bay 2 (Figure 56); these were 
indicative of extensive repair of the structure. Interior repair was also 
indicated by replacement postholes on the central wall on the Bay 2 side. 
The rows of postholes in Bay 1 and its lack of interior replacement posts, 
coupled with the lack of such rows and the presence of interior replacement 
posts in Bay 2, indicated that Bay 1 may have had a raised wood floor and Bay 
2 did not. This assumes that interior replacement posts could only be placed 
if there was no wood floor in the way. The opening in the southwest wall 
showed a cluster of five or possibly six postholes in a semicircular align- 
ment. These may have represented a porch or shed addition. There was no 
indication of a hearth in this area. An artifact pattern pertaining exclu- 
sively to this structure and the other structures in the stripped area could 
not be developed because of the nature of the impacts prior to mitigation. 



160 



uo O 

— I — r O 



O 

.on >, 

*\J CO 



_ — ^o 



fc\l aa 



o 

o 



o 


on 




oo 


ao 




— * 




O 



C\J Q3 



— — i o 



O 



a 

'-O 



>o 



o 
oo 



a. 



, — > 











■r- >> 










O "3 










</l .— 










Q. u 










O 










•>-> -a 








t- 


01 








ai 


>» s- 








■,- a> 


T3 








o. *■> 


= >, 








<o 


■O r— 








.* T3 


</t OJ 


on 


O 


o 


U 


E_ 


co 


IT) 


CM 


1_ — 




CM 


CM 


CM 


II S. 
01 

ao <j 

- c 
0) -o 

•— 4> 

>*- 


x c 

•— 4) 

£ 

II o 
<v o 

pa. = 


- 


- 


- 


«-> o 


u "~ 


00 


O 


o 


</» 


II 


e> 


Lf> 


~ 


o </l 

a.— 


4> O 



• ■>-> a. 



/a a> -•— 



o 

i> — 



ai -— — 



(J 


^— -"~ 


- — ■ 


o 


ai 


« 


"O M 






0) o 


a. 


L. ■ — 


>^-o 


*j 


•— c 


oo 


3 


(. o 


*w 


o « 




*J — 


ai .= 


■»-> 


I *J 


v. 


O f— 


> •!-> 


c 


1/1 


o 


3 •>- 


o a> 


01 


■i-J M- 


o 


i- Ll. 


o s: 


•^ 


wl O 


pM* 


+J 


01 


1_ 


O 


Ll. 


OO 


ce 


o 


Co 






<TJ ■•- 



* o ^ o o 



£ 


**- 


41 


u 


T3 


>> S 


B 


j_J XI 


4) 


cu 


X '_ 


'_ II 


i_ 


4) 


a> 


■u 


4-» II 


a. 
ii ••- 


it 


0) <_J 


L. 




41 O 


<_} <-> 


h- 


oo at 


C3 OO 



161 



ft 






e 

x 

— o 



— < > 



^O CO 



L. 

^2 



^- 


01 




41 


d 


</! 


C\J 




C\J 





CL t — » 



3 


o 


01 






<•> 


3 


i- <U 


•^ 


<_) u 


£ 






W *M 


•"-> Q. 


4-1 


4-1 


*-> 


-C 


1_ 


(J <— 


i/l >, 


u 


C TJ 


Ol 


■w 


o 


3 -f- 


C 1— 


c 


TS O 


c 


T3 


o 


1- u_ 


o 


3 


Ol 


01 


■p* 


r— 


*J 



>%-o 



01 4-> — 













■o 












Ol 












e 












■© o 












c ■— 






-C 


<c 


«* 


n *j 






Z 


z 


= 


ler 

e 

dy topsoil 

red clay 

hand excava 






















<a.*-> c >, 












«3 13 , — :l 












JCO « 11 












U 1- > 






< 


< 


< 


<*- u "O "«" i^ 






Z 


z 




hole, B = br 
mean cerami 
ay, M = mixe 
ostly or ent 
ripping, exc 






C 


c 


c 


». , — e ^-> 






X 


X 


X 


*-> o u </> 






© 


o 


e 


VI II 






c 


c 


C 


O </l >> — 






-i 


Ji 


^ 


a.-"- oi <_j « 






c 


c 


C 


Ul L tj 






3 


3 


3 


ench, P = 
xt for ba 
red and g 
rey clay, 
= mechani 




O) 








UD CT> 


0) 


u 

c 








4-> 4-> II Q. 
II i- 


o 


•fl 


Ol 


E 


E 


"DO t- 


z 


«J 


O) 


3 


3 


11 CJ O *■> 




1/1 


T) 


G 




K^aoon 


+-> 


■^ 


L. 




•S 


(J^ 


Q 


01 


X 


C 


o 




> 


T3 


— 


— < CM CI Tf 


a. 




<t 


£ 


X 



162 



•o 


-o 


■a 


■o 


c 


= >% 


c 


c 


<o 


ffl ^ 


T3 


H3 


t/> 




1/1 


i/> 


>1 


^ 


^ 


•>> 


V 


Ol T3 


0) 


9) 


S. 


t- <U 


u 


S. 


Ol 


Ol 4. 


oi 


a> 



V; in 



fl « 3tJ 



O 
O 



> I — 

^J «C Q. 



.— ^ 



*J ■!-> O 



T3 

C 
O 



T3 

C 

o 



73 
C 

o 



T3 

c 

o 



.»- c- 



a. — 
a. 



— — >, 



-H cvi 



oi — ■ > — « 



163 







■o 










«} 


13 


>. 






c 


c 










l/» 








T3 


o 








u 


>* 








■g 


t_ 


r— 






>* 


0) 


01 


■o 


"O 


c 


-O 


u 


"O 


"8 


■a 


>< 


<- 


c 


C 


T3 






c 


c 


■n 


en 


"3 


V 


i/» 


-a 


■o 


'O 


•fl 


■o 


,— 




tn 


t/l 


C 


a> 


11 


trt 


\A 


i/> 


u 


01 


>» 


>< 


I 


4-J 




>> 


>% 


*> 


>» 


••— 


<U 


01 


o 


*j 


■o 


ai 


o> 


JC 


<u 


J. 


i. 


!_ 


s. 


o 


c 


(. 


l_ 


l/> 


l. 


<o 


o> 


o» 


J3 


£ 


« 


tn 


a> 


« 


CTl 


> 



ho <" u 

oi <-> ■— 

*- !_> O 

•— o 

I** <"• V. U. 

*/» t. a) 

,uj < a. 



CO 

cr> 



O 
O 



■o o 

■1) >/> 

£ QC JD 

*-> 3 

Q. O «/) 



O 
0J 

3 
C 



ij_ ii ll. 



<L> 



01 



3 <U 

*-> jO 

u e 

3 3 

i. C 

4-1 C 

on rs 



i — <_><_> 



•— 10 — I 



164 



« ! 



■-o 



"3 »« 

in 'Z 






T I I 



o o o 



on o 



— i O O 



— O O 



— • — i cvj 



O 
O 



CVJ — I 



OJ — — < 



3 
O 

13 



•*- (/» 



U W 4-> 



O i- 



S3 



SH 



o 
o 



o 
o 



oo — • 



«3 



165 



uj \n 



i i i i 



O 

o 



ki 



o o o o 



CM O O cvi 



o 

o 



cs 





o 


tl 


■- 


o 


o 


c 




01 


-t-> 




u 


c 


•c 


0J 


u- 



o o o 



— 1/1 



4J 


■w 


w 




<rm 


C 


jB 


•■— 


■*M 




-C 


o 


u 


JC 


c 


i/> 


•u 


(/> 


•u 


u 


1_ 


e= 


o 


L. 




b 


3 


C 




01 


b* 


< 


U. 


< 


O 


a. 






166 






o 


ro 




U3 


n 


i o 


00 


<NJ 


: o 


— ' 




i . q 



is 

s 



!s 



u> 



P 



ICM 



in — • 



a 



CO 
CO 



i. 



!o 



L. 

■< 



I" 






|CM 



o 
o 



167 



p.. 



o 
co 



P 



o 
o 



o 

o 



m 









o 
o 



f 



I - 



o 









ai 








a. 
















Q. 




o> 


r— 






c 


<o 


O 




•1~ 


c 


u 




.c 


o 


<J 


t/> 


4-> 


Lrt 


« 


c 


o 


i. 


n 


(. 




01 


o 


«t 


<_> 


a. 


H- 



— 1/1 



168 



o 
o 



r»» OS O O O O — « CM 



o 
o 



o 

o 



o 



o-*oo — ooo 



a. 

3 

o 
w 





« 










<u 






s. 










a. 






3 










fm 


Ul 




*j 


41 








Q. 


a> 




u 


L. 




en 








e 


4> 


3 




e 


« 


O 


■w 


<u 


*-> 


w 






c 


u 




.c 


-— 


-— 




JZ 


o 


o 


> 


<j 


-C 


e 


<si 


«-> 


1/1 


"3 




*■> 


u 


s_ 


S 


o 


s. 


J3 


•u 


^» 


I. 


a 


t_ 


.— 


01 


o 


u 


^ 


< 


u. 


< 


<_) 


a. 


►— 


< 



169 



O 




CQ 

un cm 



CM 

t^3 :*£ 

IT) CQ 

CO 

LU CO 

■=> CU 
CO -M 



GJ 

S. Cfl 

■m ^r 

u CM 

zs 

■<- +J c 



o 



o 



CM 
UJ 

<r 
o 

3 

cr 



o 



o 







170 



This structure, like the majority of structures at 388K245, was better 
constructed than those of Site 38BK76. This is seen in the care with which 
the trenches of Structure 245B were dug, the careful alignment of the 
trenches, the use of hewn or sawed posts in some instances, the integrated 
end and central walls, the similarity of orientation of the structures, and 
the consistency of the measurements of Structures 245B and 245D, the large 
two bay structures, as compared to the lack of consistency of structures at 
38BK76. 

Excavation of Structure 245B included all isolated postholes, the replacement 
postholes along the southeast wall, two cross-sections of the long trenches, 
the northeast trench, and excavation of two to three inches of the remaining 
trenches in order to examine them for postmolds (Figure 54). Time spent at 
Structure 245C prevented complete excavation of the trenches. The fill of 
the trenches at 245B was remarkable for the mixture of red clay subsoil (as 
at 38BK76) and a dark organic, elastic gray clay perhaps obtained from the 
San tee River floodplain. 

Chemical analysis of the trenches at Structure 245B showed that the trench 
fill fell within the significance limits for phosphate, nitrogen, and carbon 
(Appendix C). It should be noted that an extremely acid pH value very 
different from that of the natural soils and feature fill at both 38Bk76 and 
388k76 characterized the entire site. This indicated that bone preservation 
should be very poor at 38BK245. However, there was much more bone at 38BK245 
than either 38BK76 or 38BK76, despite the poor preservation potential at the 
former. Indeed, 294.1 grams of bone were collected from the northeast wall 
at Structure 245B and from postholes I and M, which represented more than the 
total of all features at Site 38BK75 or 38BK76. Of this bone, 286.64 grams 
were cow and the remainder pig. In comparison, Site 38BK75 had a total bone 
weight of 230.8 grams and 38BK76 had a total of 248.16 grams. 

To the northwest of Structure 245B was a posthole structure, Structure 245G. 
This structure consisted of two parallel lines of very neatly excavated 
square postholes, several of which had round postmolds. The wide spacing of 
the postholes and particularly the fact that the northeast and southwest ends 
had no intermediate postholes led to the conclusion that this structure 
probably served a different function than the other structures at Site 
38BK245. Post placement seemed to indicate an open-sided structure, perhaps 
a covered work area. 

A small feature, probably a trash pit, was found on the western corner of 
Structure 245G and may have been associated with the structure. Feature F2 
extended .45 foot into subsoil and was remarkable for the extremely high 
frequency (nearly 15 percent) of pipe parts (Table 18). This suggested that 
the feature was near an area of leisure time activity. 

Structure 245H (Figure 57), located south of Structures 245B and 245G, was 
outstanding for several reasons. It was the largest structure encountered 
during the project, 986.6 square feet, or 662.6 square feet larger than the 
next largest building unit at Structure 76B2 (see below for a discussion of 
building units); it had two entrances facing the same direction (northwest), 
along the long side of the structure; the posts used in the construction of 



171 



the building measured 3 by 3 to 3 by 4 inches (the approximate equivalent of 
modern 4 by 4 posts) and were hewn or sawed; the bottom ends of the posts had 
been burned before placement to preserve the wood (yellow pine); and the 
distance between posts and the alignment of the posts in the trenches was 
very regular. Because it was so unlike any of the other structures in size, 
layout, and construction, it probably fulfilled a function different from 
those structures interpreted as cabins or sheds. A line of posts stretched 
from the southeast exterior wall almost to the northwestern exterior wall on 
the northern end of the structure (Figure 57). Apparently, this line of 
posts either separated the northern end of the structure in order to form a 
small room, or provided support for a loft, or both. If a small room was 
formed, it had a separate entrance. There may have been a line of posts 
running southeast to northwest across the middle of the structure, but this 
line was not as clear as the first. There were also other scattered posts 
inside the structure, but their alignment was not clear. It should be noted 
that in most of the structures excavated at Yaughan and Curriboo, there were 
few interior posts and rarely were they as clearly aligned as in the first 
row mentioned above. The positions of the scattered posts, so unlike the 
regularity apparent in the exterior walls, argues against a wooden floor. 

This structure could be described as a large, well-made structure with a dirt 
floor, two large doors on the same wall, and a loft or small room on the 
northern end. The structure was sufficiently unlike the other structures to 
suggest that it had a different function. Three other pieces of information 
provided insight into the structure's function. Feature 245F59 may have 
represented a manure pile box and seemed to be associated with the structure 
(see discussion below). Structure 245C, potentially an administrative of- 
fice, was located 1200 feet to the north and probably represented the north- 
ern edge of the plantation settlement. Between Structures 245H and 245C were 
slave houses or cabins, and between the slave cabins and the office were 
concentrations of brick so thick that a plow could not penetrate the soil 
(Joseph Cooper, personal communication) and where extensive brick and fea- 
tures suggest the owner's house was found during excavation of the borrow pit 
(unnamed construction worker, personal communication). This information 
would suggest that Structure 245H was not the main plantation house (it had a 
dirt floor and it was too far from the office and too close to the slaves), 
but rather a special function structure located in the slave quarter. It was 
too large for a corn crib or smoke house so it was concluded that it was 
probably a barn. 

Feature F59, mentioned above, extended only .1 or .2 foot into subsoil on the 
northern side to less than .1 foot on the southern side. It had been heavily 
impacted by topsoil stripping conducted before mitigation. Unfortunately, 
there was so little soil in the feature that it had been screened before a 
soil sample was taken . No chemical tests could be run and no bone was recov- 
ered. The location and orientation of the feature clearly associated it with 
the barn, but its function was unclear. Because of its shape, location near 
and orientation with a possible barn, and the dark organic quality of its 
fill, it may have represented a manure pile. 




FIGURE 57 
Site 38BK245 
Structure 245H 



173 



Feature F12, a clay extraction/trash pit, was located equidistant to the 
northwest of Structure 245H and to the west of 245G (Figure 27). The fea- 
ture's association was, therfore, unclear. The feature extended 1.3 feet 
into subsoil and contained approximately 32.4 artifacts per cubic foot. 
Level 1 (Figure 58) was a dark humic fill which represented the last filling 
episode. Level 2 was water-laid sand occurring only near the sides of the 
pit, and Level 3 was gray to black organic fill which represented the first 
filling episode. Chemical tests for all levels were within the normal range 
and were not those associated with good preservation (pH ranges were from 3.6 
to 3.9). The feature contained 14.94 grams of bone, however. Of this, 9.49 
grams were burned, which explained its preservation. Two unidentified carpal 
bones showed evidence of having been sawed during butchering. Animals known 
to be represented were pig and catfish. 

The mean ceramic dates for Levels 1 and 3 were 1764.83 and 1750.33, res- 
pectively. The difference between them was 14.5 years, implying that the pit 
remained open for that length of time between the first and second disposal 
episodes. The amount of water-laid sand did not bear this out, and the dif- 
ference in dates may have been linked to sample size, since Level 3 only had 
six datable sherds and Level 1 had 54. 

This feature contained a wide variety of artifacts outside of the usual 
Kitchen, Architecture, and Pipe Groups, the groups where most material was 
found in features at 388k75 and 38Bk76. Level 1 contained two gun flints, 
one brass ring (possible harness part), one key, one sickle blade fragment, 
one harness buckle, and a hook-like piece of iron. Level 2 contained a gun 
flint and a bale seal with "W D" stamped on it. Level 3 contained a fragment 
of a half-round rasp and a brass furniture tack. A slate gorget of indige- 
nous American manufacture also came from the feature, but was unfortunately 
not recorded by level. This wide variety of artifacts illustrated activities 
such as hunting, rice or indigo harvesting, care of horses or cattle, com- 
merce, carpentry, artifact collecting, and implied others. Even the cellars 
at 38BK245 did not represent so many different activities. 

Without block excavation of the surrounding structures (245B and 245H), a 
complete analysis of this feature, its function, and associations could not 
be made. It is evident, however, that the feature remained open or was used 
for a considerable length of time and was the focal point of disposal from 
various activity areas within the plantation. 

Structure 245D was virtually identical to Structure 245B, except for its lack 
of an end trench on the northeast wall (Figure 59). No excavation of the 
trench fill was conducted, but stripping had been enough to expose the post- 
mold pattern in the trenches. This pattern proved to be clearer than that in 
245B, despite hand excavation of the latter. Interior postholes indicate 
that there may have been a raised floor in Bay 1 of this structure, similar 
to Structure 245B. Replacement posts along the interior of the southeast 
wall of Bay 2 and along the exterior of the northwest wall indicated exten- 
sive repair work not apparent 1n Bay 1. 

Feature F62 (Figure 27), a clay extraction/ trash pit, which seemed to be 
associated with Structure 2450, extended .7 foot into subsoil. Along the 
northern edge was a lens of light water-laid sand, indicating the pit was 



174 



x 

O 

<n 

z 
o 



UJ 



o 
<r 

a. 

H 

cn 

UJ 



< 
UJ 



Qu. 

<<r 

(TO 




MP 
Kjgpta 



'?*$& 



mmm 



SrfwS 




£s£38LS SAW*! 



Hi* 






tffl» 







en 

< 

UJ 

o 

z 

a 

< 



UJ 

Gl 
O 

ac 
a. 

z 

H- 

O 
<n 

i 

z 

Z 

o 





Ki^s^'jSEj^' fibs 



£&& 



§*y 






i35£ 






^i™?,^ 

>>«&:*.- 



rt- -*'£'.' 



in 






►^'■^ 



SSSE 






SS'riftl 

: *S& 















yjs&iii 



CM 



UJ 
UJ 

u. 





If) 






<a- 


CVJ 




CM 


i— ( 


cc 


^ 


Li_ 


un 


CD 






oo 


<U 


LJ 


m 


s_ 


C^ 




3 


3 


O) 


+J 


CT 


-m 


CO 


h— ^ 


•r- 


OJ 



Li- OO Ll. 



UJ 

-I 
< 

o 

V) 



o 
o 

2 

o 

o 



Eg 



o — 

ou. 



e 
o 
en 






at 

o 



£ 
z 

J* 
u 

o 

CD 

o 



o 



o 

o 



p 











175 



o 



a 





3 



% 



a 



& 



O 

O 




o 

(VI 




o 



^ o 




a 
un 

CM 
O) 

s_ 

3 
4-> 

O 

ID 3 

^r s_ 

LT> CO 
CO «*- 

lu on o 

OS 

C3 +■> (T3 

LLOl Q. 



O 



o 



a 



176 



open for a period before it was filled in. The approximate number of 
artifacts per cubic foot was 30. Chemical soil analysis of the feature 
showed that it was within the normal range of the features at 38BK245. Cow 
bones represented 52.89 grams, or the total faunal assemblage. 

Feature F64 (Figure 27), a trash pit also apparently associated with 
Structure 245D, was a shallow amorphous layer of gray sand which did not 
clearly extend into the subsoil. It was treated as a feature because of a 
hoe exposed on the surface. In final analysis, this may have been remnant 
topsoil which was not completely stripped away. 

Structures 245E and 245F were heavily impacted, not only by stripping, but by 
the parking of heavy equipment in the area before mitigation (Figure 59). 
The northeastern end of the northwestern trench at Structure 245E had been 
completely destroyed and the remaining trenches had been impacted almost to 
their bottoms, causing some of the irregularity in their outlines. As with 
the other structures in the stripped area at 38BK245, no artifacts could be 
definitely associated with the structure. An alignment of postholes along 
the exterior of the southeastern wall had no clear functional explanation. A 
presumed second trench and an unknown portion of the remaining trench at 
Structure 245F had been destroyed before mitigation. This structure could 
not be used for the pruposes of architectural comparison. 

Structure 245A (Figure 60), a cellar, was quartered and the western and 
eastern quadrants were excavated by natural layers to the sand floor at the 
bottom of the cellar. After completion of profiles through the cellar, the 
north and south quadrants were excavated. There were essentially three 
levels within the cellar fill. Level 1 was red clay washed into the cellar 
hole after the building over the cellar (of which there was virtually no 
trace) had disappeared. Level 2 was a dark organic clayey soil with lenses 
of ash and charcoal which may have represented the burning of the structure. 
Level 3 was the one to two inch thick coarse sand floor, mixed with the red 
clay subsoil at their interface. Layering of Levels 1 and 2 could be noted 
along the northwest wall (Figure 60); this may have represented the period 
from abandonment to destruction of the building above the cellar. The cellar 
extended at least 2.2 feet into subsoil, but had been so badly damaged at the 
top prior to mitigation that its maximum depth could not even be guessed at. 
As can be seen in Figure 60, the eastern corner of the cellar was completely 
destroyed. 

As the floor was approached during excavation, all artifacts were mapped 
in-situ . The artifacts represented in Figures 60 and 61 are only those which 
were lying directly on the sand floor. Feature 245AF1 in the north quadrant 
was a pile of construction debris and trash which included a restorable 
colono pedestal pot, a barrel hoop, stones, and bricks. Another restorable 
colono pot was found in Level 2. No bone was found in the cellar. The 
addition of bricks from the floor to the artifact pattern would result in a 
higher Architecture Group percentage, which may represent the cellar fill 
more accurately than the pattern given in Table 18. A revised artifact pat- 
tern for the cellar which includes the bricks from the floor is presented in 
Table 19. Even with the addition of 32 more architecture artifacts, the over- 
all pattern still does not fit into the revised Carolina Artifact Pattern, 



177 



8' 



'-ggewo 

□ Brick 
O Rock 
f Noil 




Light Rtd Clay 



□ 



Packed Sand 



hi"*' | G'oy Cloy «>tti 
>'- *''■'■* Charcoal 



, Ash Ltn»» 



:" U R«d Cloy Subsoil 



CHARRED WOOO 



FIGURE 60 
Site 38BK245 
Structure 245A 
Cellar Plan 
and Profile 



178 




STRUCTURE 245 A 
FEATURE 245 A Fl 
LOOKING NORTHWEST 




STRUCTURE 245 K 
LOOKING SOUTH 



FIGURE 61 
Photos of 
Structures 
245A and 245K 



although it comes closer. 



TABLE 19 



Revised Cellar Artifact Pattern 

(includes bricks from floor) 

Structure 245A 



179 



T 



T 



Kitchen 

Architecture 

Furniture 

Arms 

Clothing 

Personal 

Tobacco Pipes 

Activities 



310 
124 
1 
1 
2 

11 
3 



68.58 

27.43 

.22 

.22 

.44 

2.43 
.66 



TOTAL 



452 



99.98 



Features 3 through 10 were hypothesized to represent irrigation channels or 
ditches running across the site, usually northeast to southwest. A five- foot 
section was taken from the trenches to recover artifacts and to obtain pro- 
files and bottom depths in order to determine the direction of water flow. 
Most trenches were relatively shallow with neatly excavated bottoms. Often 
they contained a thin lens of water-laid sand on the bottom under the dark 
organic fill. 

Feature F3 is one of eight trenches (Figure 27). This trench made a 90° turn 
to the northwest on its southwestern end and represented only a fragment of a 
more extensive trench which was destroyed before mitigation. It may have con- 
nected with Trenches F4 or F5, or both. Feature F3 flowed from the southwest 
to the northeast. 



Feature F4 was a fragment of a trench in front of Structure 245H. It flowed 
from the southwest to the northeast and did not appear to have been deliber- 
ately filled with trash as Feature F3 was. A lens of water-laid sand ap- 
peared at the bottom. 

Feature F5 (Figure 62) predates Structure 245B, whose northwest wall intruded 
into and was parallel to the trench. Water flow was again from the southwest 
to the northeast. The trench did not appear to have been deliberately re- 
filled with trash. Two lenses of water-laid sand may have indicated two 
filling episodes. 

Feature F6 (Figure 62) was the southeastern of three associated trenches. 
Water flow was from the southwest to the northeast. The artifact-bearing 
fill was underlain by a relatively thick layer of water-laid sand. Feature 
F7 was the central of the three associated trenches. Water flow was again 
from the southwest to the northeast. There were two lenses of water-laid 
sand; one was at the bottom of the trench and below a sandy red clay layer 



UJ 

X 





180 



o 

V) 

>» 
o 

o 



o 
V) 

o 



I 



o 
o 

>> 

8 
en 

* 

o 



>. 
o 



o 
O 




5 




£ 


• 













UJ 

o 

E 

x 
i- 
<r 
o 




lA 





E 




O 




+J 




+J 




o 




-— . -Q 




C — « 




o 




— JZ 




^- o 




c 




-C O) 




<J s- 




C 1— 




OJ 




S_ \0 




1— U- 




LD 




^r i O) 




CM LD S. 


ex 


^. Ll_ 3 


UD 


C3 -M 




CO OJ <o 


UJ 


co i- <D 


Q£ 


3 Ll- 


_D 


O) — ■ 


U3 


■*-> rO "O 


i-^ 


— OJ c 


u_ 


on u_ ra 




181 



and the second was between the sandy red clay and the gray organic soil. 
Feature F8 was the northwesternmost of the three associated trenches. Water 
flow would have been from the southwest to the northeast. A lens of water- 
laid sandy silt appeared on the bottom of the trench along the southeastern 
edge. This was the only feature that showed evidence of the shovels used to 
excavate the trench. There were four possible shovel cuts in the bottom made 
by a round-nose shovel which was curved in cross section. The shovel was at 
least 5-6 inches wide. 

Feature F9 was one of two parallel trenches to the west of the three asso- 
ciated trenches. The major portion of the fill was gray organic sand. Below 
this was a thick lens of white water-laid sand as in most of the previous 
trenches, but below this lens was a layer of tan coarser sand which may have 
also been water-laid. The coarser material implied a faster water flow than 
the finer water-laid sand lenses. The direction of water flow, from north- 
east to southwest, was reversed for this trench and its associated trench, 
Feature F10. Feature F10 was the northwesternmost of the two associated 
trenches. The trench fill was underlain by a thick layer of water-laid sand. 

These eight trenches were similar in many respects and probably fulfilled a 
similar primary function. This function was determined to be irrigation. 
This determination was based on the presence of one and sometimes two lenses 
of water-laid sand which implied that the trenches remained open long enough 
to collect such material, the evident care with which the bottoms were dug, 
the consistency with which the slope of the bottom was maintained, the simi- 
larity of the fill in the trenches, the relative lack of artifacts (except at 
F3), the length of the trenches, and the lack of postmolds in any of the 
trenches. 

All of the trenches except F9 and F10 ran downhill from the southwest to the 
northeast (Figure 27). Features F9 and F10 ran in the opposite direction. 
The shallowest trench in absolute elevation below an arbitrary datum was F3, 
followed by F8, F7, F10, F4, F5, and F6; the deepest was F9. Such variation 
from trench to trench and from associated group of trenches to associated 
group argued against a single source of water supply. It had been hoped, in 
the field, to show how the water may have circulated from a water source at a 
higher elevation. This was not possible with the data at hand. The hypothe- 
sis that water zigzagged through the trenches, reversing direction in each 
trench, has also proved false. Apparently, there were trenches outside the 
project area which could answer questions of water source and circulation, 
but some of these have undoubtedly been completely destroyed, and the others 
are beyond the physical boundaries of this project. 

The trenches appeared to antedate at least one and possibly more of the struc- 
tures. What the trenches were irrigating is undeterminable at this time, 
although vineyards were suggested. Associated with the trenches were irreg- 
ularly spaced postholes, which intruded into the trenches and postholes near 
the trenches. The purpose of these could not be determined. 



182 



Along Trench F10 on its northwestern side were four trash pits evenly spaced 
from the trench. These and one other pit near an unexcavated trench are 
discussed here because they appeared to be aligned and associated with the 
trenches. Feature Fl, a trash pit, extended .37 foot into subsoil and had a 
rounded bottom. The pit contained 5.12 grams of oyster shell. Feature F60, 
a trash pit, extended .31 foot into subsoil and had a rounded bottom. Only 
the western half of the feature was excavated. Feature F61, a trash pit, 
extended .23 foot into subsoil. Only the west half was excavated. The 
similarity between these features and their shallowness suggested a line of 
decorative plants along Trench F10, although this is purely speculation. The 
shallowness of the features did seem to argue against their function as clay 
extraction pits. 

Feature F63, a clay extraction/trash pit (Figure 63), extended 1.71 feet into 
subsoil and presented a more complicated stratigraphy than the other pits 
associated with Trench F10. The south side of the feature had a mixture of 
dark brown sand with lighter sand and red clay mottling, and sloped from near 
the top to the center of the feature. This was overlain by dark brown sand 
sloping in the same direction. On the top was the culture-bearing dark gray 
organic sand as in most other features. This layer also contained 7.49 grams 
of cow bone. The mixing of the bottom layer may have been, the result of back- 
filling the hole soon after it was dug with a mixture of topsoil and subsoil. 
The brown sand layer was similar to existing topsoil in nearby fields and the 
top layer was a mixture of this topsoil and organic debris. The layers could 
be explained if the hole was dug, the clay was removed, the clay/topsoil in- 
terface was replaced, the unused topsoil was then replaced and the remaining 
space was filled with topsoil and trash at a later time. This pattern may 
have been the clearest example of clay extraction at any of the sites, with 
the exception of Features F30, F31, and F32 at 38BK75. 

Feature F65, a clay extraction/trash pit, was in association with an unexca- 
vated irrigation trench, just as the previously discussed pits appeared to be 
associated with Trench F10 (Figure 27). The feature extended .84 foot into 
subsoil, and thus was more comparable with F63 than with the other pits. It 
was also similar stratigraphically to F63, although soil changes were less 
dramatic and more a matter of degree within the gray organic layer common to 
the other features at 38BK245. All levels produced artifacts and layering 
appeared to be due to irregular episodes of filling. The most remarkable at- 
tribute of this feature was the elevated amount of Architecture Group arti- 
facts (Table 18), which included 11 window glass sherds. This feature may 
have represented debris from Structure 245D. The feature also contained 2.87 
grams of burned bone. 

Structure 245K, a brick clamp in the large borrow pit to the north of the 
previously discussed structures and features (Figure 27), had been heavily 
impacted by excavation of a soil borrow area entirely around the feature 
which left it on a pedestal (Figure 61). The surface of the clamp was also 
scraped by a tractor, as evidenced by extensive tractor tread marks into the 
exposed red clay subsoil (Figure 64). Later, the clamp was covered with 
black plastic and sterile sand to preserve what was left until mitigation. 
However, as a result of extensive damage by the tractor and the shallow 



183 




.U'jM.'MJJ' 



Dark Brown Sand 



Vkv'j^'A 8lacK Organic Fill Mottled with 
Clay and Cultural Debris 



Dark Brown Sand Mottled with Light 
i&£fl Yellow Sand and Clay Inclusions 



GRADED 
SURFACE 




I 

SCALE IN FEET 



WEST PROFILE 



FIGURE 63 
Site 38BK245 
Feature F63 
Plan and Profile 



184 







^ 






LO 




LO 


■^r 




«3" 


CM 




CM 




^r 


^. 


<U 


U3 


CO 


s. 




co 


3 


LjJ 


o-> 


■M 


cc 




'O 


X) 


0) 


3 


u 


— 


- 



OJ 



■m a. 





U. I/O iS) ^L 



< 

E 

3 
O 

a 







UJ 








> 






K 


< 


• 




UJ 


X 


a 




UJ 




UJ 




u. 




h- 


T 


Z 

UJ 


CO 

Z 


< 
> 
< 




-J 


X> 


U 




< 




X 




cj 




UJ 


C\J 


CO 


a 

UJ 








a 


z 


— 




< 


UJ 






X 


UJ 


n 




CO 


CD 



185 



nature of such features, only a charcoal stain and scattered, broken brick re- 
mained to be excavated. 

The area was trenched from north to south to provide a profile of the deepest 
part of the presumed intact portion of the clamp. The remaining area was di- 
vided into quadrants and excavated. Three charcoal concentrations with as- 
sociated brick concentrations were delineated. Several isolated brick con- 
centrations were probably the result of the tractor disturbance. Feature 
245KF1 (Figure 64) was a trough or ditch with a thick concentration of 
charcoal and brick. Its irregular depth and width precluded it from being a 
functional part of the clamp. No in-situ bricks from the clamp which might 
have provided insight into its size or form were found. 

The primary result of the clamp excavation was the development of the arti- 
fact pattern presented in Table 18 (high architecture and activities per- 
centages and an overall low artifact count) and a large quantity of brick 
fragments which could be used for comparative data at Structure 245C, the 
brick piered structure. All brick fragments with at least two measureable 
dimensions were returned to the lab for measurement. One of these had an 
"S-..." (Samuel Cordes?) written on it with a finger while the clay was still 
soft. Brick measurements and comparative data from Structure 245C are dis- 
cussed below in detail. 

Structure 245C, located north of the borrow pit and the other structures and 
features already discussed, was excavated in 10 foot by 10 foot squares 
(Figure 65). Brick rubble was cleared and a large sample of complete bricks 
from the rubble were boxed and returned to the lab for measurements. The 
first layer of soil was mixed with rubble, dark leaf mold, shell from the 
tabby mortar, and topsoil . Below this, the soil was a mixture of brown top- 
soil with small amounts of shell and occasionally brick rubble. This layer 
rested on red clay subsoil (Figure 66). Structure 245C actually included two 
structures, 245C^ and 245C2« Structure 245C^ is discussed first. 

To the northeast of the chimney (Figures 65 and 67), the original ground sur- 
face dropped dramatically to 4.0 feet below ground surface (Figures 66 and 
68). The base of the chimney rested on red clay at this level. The red clay 
extended further to the northeast from the chimney base and continued down to 
3.2 feet before meeting the east wall of the excavated block. This fill 
northeast of the chimney consisted of several layers, and all but the top .8 
foot was laid down soon after construction of the chimney base. The lowest 
level was an ash lens. This was covered with mottled red and gray clay which 
contained construction artifacts. The next layer did not extend far from the 
chimney, where it was thickest, and feathered out approximately three feet 
from the chimney. This layer was a mixture of sand and gray clay* A gray 
sand mortar and brick rubble lens was superimposed on the sand and gray clay, 
which possibly dated from construction of the chimney. All of these layers 
were then covered by a mottled red clay and tan sand layer with few artifacts 
and later by topsoil and brick rubble. 

Outside of the excavation block and yet again further to the northeast, there 
was a slight depression in the natural surface running southeast to 



186 




in 
c\j 

U3 CQ 

cc 
an 

=> CD 
<L3 4-> 



IT) 
<T 
CM 

o 

1- 

U 

3 
S_ 



C 

O 



o 

o 



U- (/I (/I Cfl 



3 

UJ - 




•I — I 



187 



P 












c 










o 










c_> m- 










LO •<-> 








IT) 


^r <T3 








>sr 


CM > 








CM 


<T3 








U3 i^ 


O) <J 








wo ao 


1_ X 








00 


■=, i— j ai 








UJ PO 


4-> i — 








d 


u -s<: •■- 








3 O) 


3 u <+- 








O +J 


s- o o 








»— * <^ 


+J -— i- 






A 


U_ t/1 


i/ioaa. 
















i- 






a 




=1 






s 




gM 






* 




3° 






% 


e 






* 


s 


« 






£ 


ID 


•o •• 






li 


e 
a 


• < 

If 






° > 




3 5 






a* 


■9 
• 


•»" 






5 3 


r 

O 

2 









D 



I 



3 
o 

& 

s 

a. 





is 

I 

O C 


S 

I 

o 

org 


1 

a 
o 

»» 
o 

ll 

1? 


1 

o 
in 

3 

3 

9 


so 


31^ 


3o 


J 



□ 



188 




STRUCTURE 245 C 
LOOKING SOUTHEAST 




CHIMNEY AT STRUCTURE 245 C 
LOOKING SOUTHWEST 



FIGURE 67 
Photos of 
Structure 245C 
and Chimney 



189 



a 



UJ 
UJ 



C\J Z 




T3 

«3" CM 3 f— 

CM O ■«- 

oo *£ 0) u- <+- 

CO 3 >> S- 

uj co -t-> a) ca. 

=i <U 3 E i— 

CJ +J S_ •<- r— 

i— i •<- 4-> .C -t- 




e 
o 




"2 






>» 




>» 


o 


•o 


a 


o 


e 
o 

(A 




>» 
O 




■o ° 
• </» 


w 

o 


** 5 








Mott 
Gray 


§ 

-J 


Mott 
Gray 


< 












m 








'.'■: ''.:' 


Ik. 1 








'. - r 


ijssa 











M 

2 3 



o 

o 



-S<0 



W.S ^ 



a <- 

OCD 



2^ 



fi^a. 



s 



u 

w 

fUl 



e 
o 
en 

>» 
a 

w 

O 



£s 



190 



northwest; this may have reflected the originally lower ground level on that 
side of the building before the soil layers above were put into place. Since 
the ground may have originally sloped down and away from the structure on the 
northeast, only minor excavation of the existing slope was required to pro- 
vide for the chimney foundation. A gray ashy hearth pad was evident to the 
southwest of the chimney, and remains of the hearth could be seen on the side 
of the chimney facing the hearth pad (Figure 65). The hearth pad was .15 to 
.20 foot thick next to the chimney and feathered out from there. 

Of the 22 seeds recovered from the fire box, two were rice and two maize. 
Only Feature F15 at Site 38BK76 had more rice. This was also the only fea- 
ture at Structure 245C to have faunal remains, other than bivalve shell. 
There was .02 gram of land snail and .64 gram of opossum. However, these 
could have been deposited as naturally occurring specimens in the fill dirt. 

The building itself (Figure 67) was built upon at least nine and probably ten 
brick piers (the north corner is missing). The piers were constructed of 
four bricks per course and held together with tabby like mortar. Piers P3, 
P5, P6, P8, and P10 (Figure 65) had evidence of two courses of brick, and all 
of the piers may have been higher originally. The condition of the top 
layers of the hearth precluded estimating floor height (Figure 67). It was 
evident from the central supports (Piers P5, P6, and P8) that the structure 
had an elevated wood floor. 

An attempt to locate walls which contained windows failed. A total of only 
52 window glass fragments were recovered. Twenty-three of them were found in 
units either completely or more than halfway under the house. The heaviest 
concentration was at the interior northern corner of the structure, which 
possibly indicated a window on the northwest or northeast wall between the 
corner and the chimney. It is possible that undamaged window panes were 
salvaged and removed from the site or that the structure had few glazed 
wi ndows . 

Structure 245C2 was directly below Structure 245Ci» In fact, the piers of 
Structure 245Ci rested on and were aligned with the trenches of Structure 
245C2* Even though it was possible to obtain the width of the trench struc- 
ture, its length extended beyond the limits of the block. On the last day of 
fieldwork, an attempt was made to define the limits of the southwest trench 
with unstructured excavations. The dotted lines on Figure 66 represent the 
results of this work and give the best estimate of the remainder of the 
structure. The southern end of the southwest trench showed a "T" intersec- 
tion which would indicate a building much larger and more complex than the 
other relatively simple trench structures. 

The key to the function of Structure 245C? was the cellar associated with it. 
The cellar was clearly associated with the trench structure (245C2), rather 
than with the brick piered structure (245Ci). This was made clear by the 
discovery that the cellar was confined within the trenches, three brick piers 
were built on the cellar fill, and the hearth pad associated with the chimney 
base of 245C also extended onto the cellar fill. 



191 



The cellar (Figures 65 and 69) was sectioned and the northeast end was 
excavated to the cellar floor. The top layer of fill was gray sand with 
large quantities of tabby mortar and shell fragments. The consistency of 
this layer and its unusual nature seemed to indicate that it was intention- 
ally placed to level the cellar at ground level. Below this, and entirely 
across the cellar, was a thick layer of mottled gray clay. A thick layer of 
water-laid sand was wedged between the gray sand and mottled gray clay layers 
on the western side of the cellar. This indicated that the mottled gray clay 
was in place and left open to the elements for a considerable length of time. 
Below the gray clay was a layer of mottled reddish-orange clay. The presence 
of the gray clay overlain on the reddish orange clay was apparently a rever- 
sal of the normal subsoil stratigraphy. Both soil types did not occur ad- 
jacent to the cellar and must have been imported. The cellar had collected 
water so that everything below the orange clay layer was almost permanently 
under water. Below the reddish orange clay in the center of the cellar was a 
layer of organic clay. Chunks of poorly preserved pine boards, which were 
thought to be flooring were encapsulated in the organic clay. Immediately 
below the organic clay was a layer of hardened pine tar. The pine tar rested 
directly on the red clay floor on the western side of the cellar and on 
orange sand mixed with the red clay subsoil on the eastern side of the cel- 
lar. The pine tar had obviously built up over a long period. Pressed into 
it were the forms of palmetto and other unidentifiable leaves, which had 
slowly been covered and preserved by dripping pine tar from above. 

Understandably, carbon was exceptionally high in the organic clay layer. The 
lack of bone (and the presence of wood), the very low overall frequency of 
artifacts, the low percentage of Kitchen Group artifacts (all colono sherds), 
the high percentage of Architecture Group artifacts (all nails), and the pine 
tar layer resting on the floor indicated a special use function for the cel- 
lar. The function of the cellar and trench structure associated with it was 
probably the processing and storage of naval stores. The odor given off by 
the pine tar was distinctly reminiscent of old wooden ships, which may indeed 
have been the final destination of the tar stored or processed in the struc- 
ture above. Such tar was used in caulking and preparing rigging on sailing 
ships, and documents indicated that Samuel Cordes had been involved in the 
production of naval stores. 

Summary of 38BK245 

The mechanically stripped area of 38BK245 produced evidence of eight struc- 
tures; these were 245A, 245B, 245D, 245E, 245F, 245G, and 245H and a possible 
structure northeast of 245B. Two were large double bayed structures and one 
was a barn. One cellar was excavated, but it and the structure above it had 
been severely damaged. Associated with these structures were trash pits, 
postholes, and sets of irrigation trenches aligned from the southwest to the 
northeast. To the north of the stripped area was a severely damaged brick 
clamp (245K), and to the north of the brick clamp were the remains of a 
chimney fall. Hand excavation of the chimney fall produced a chimney base 
and a brick piered structure (2450^) overlying a trench structure (245C2). 
Associated with the trench structure was a second cellar. The brick piered 
structure served a non-domestic , probably administrative, function on the 
plantation. The trench structure and cellar were associated with naval 
stores production. 



192 



o 

CO 

V 

o» 

c 
o 

w 

o 



V 

c 
o 



•a 



O « 



o 



CO <» 



?! 

o to 

pa 

0.» 



o 



CO 

>» 

o 










':,.*_ 



o 



c 
a 



<u 

c 




"-•v. 



'2' 



0< 



^ilir 









Pef 



••• ^: 



■ 0: *-2°-£'* 

?^^ 

o^c^o r 

iig.--N*0 



*>v. 



:>.£<* 

: <«& 



,.t ':'• 



W C ? 






•\ 







o 


a; 






LTJ 


»^ 




un 


^r 


• ^~ 




W 


C\J 


4- 




CM 




O 


Ci 


*<£ 


0) 


i. 


10 


CQ 


s_ 


c_ 




cc 


3 




U_l 


CO 


■Ul 


s_ 


e£ 




U 


03 


ZD 


aj 


3 


,— 


es 


■M 


S. 


, — 


^* 


•f— 


4-) 


O) 


Lj_ 


<y0 


oo 


u 



UJ 

UJ 

u. 



< 
u 



193 



Slave Architecture 

The preceding sections have briefly described 29 structures and 58 associated 
features. Hundreds of features, including postholes and trenches, were 
mapped and recorded during the project. The structures and features alone 
could be analyzed from a variety of viewpoints, and hopefully will be in the 
future. The controlling hypotheses of this project have determined our ap- 
proach to the synthesis of the structural data. Summarized, our hypotheses 
concerning the structures were: (1) that the earliest foundation form at 
Yaughan Plantation (38BK75 and 38BK76) was trench construction, which Wa 
later superceded by posthole construction; (2) this sequence represented 
acculturation of the slave population; and (3) at Curriboo Plantation 
(38BK245) trenches were earliest, followed by postholes, and that both were 
superceded by brick pier construction for major plantation outbuildings. Our 
conclusions are presented here. 

The two primary superstructures were, in our opinion, frame and mud wall (see 
Appendix D). Each to a certain extent determined roof type and the presence 
or absence of architectural accessories such as windows and fireplaces. Wat- 
tle and daub was considered as a superstructure, but was rejected. Wattle 
and daub implied vertical posts with saplings or brush woven between them, 
covered with mud. If a trench was dug to place the posts, the wattle extend- 
ed into the trench to provide a seal around the bottom of the wall. If such 
a seal was not wanted, postholes would suffice. At Yaughan and Curriboo, 
none of the trench structures showed evidence of wattle; furthermore, some of 
the postholes on posthole structures were too widely set for wattle to be 
woven effectively. Our conclusion is, therefore, that neither foundation 
type represented wattle and daub superstructures. 

The requirements for mud wall architecture were adequately met by trench con- 
struction. The trenches were not simply backfilled with unmodified topsoil 
but were filled with clay which occasionally showed evidence of being mixed 
with water before application. Trenches would preclude, to a certain extent, 
undercutting of a mud wall and provide support of lateral thrust as discussed 
in Appendix D. A stone foundation would do the same, much like adobe struc- 
tures in Mexico (personal observation 1972-1976), but building stone was in 
short supply in the Coastal Plain. The foot wide or often wider wall bases 
would be necessary to support a heavy, probably tapering wall, above. The 
rather closely set posts would give added strength to such a wall, much like 
rebars in concrete. Further, trench foundations were found at St. Genevieve, 
Missouri (Fairhurst 1974), Fort Michilimackinac (Marlesa Gray personal com- 
munication 1979 and Stone 1974), and the Gulf Coast (Wilson 1979), where the 
superstructures were hypothesized to be open beam with other materials such 
as brick or plaster filling the interstices. At the Cooper River sites, the 
space between posts was clearly not filled with brick or plaster, but rather 
clay mixed with water to a mortarlike consistency. Evidence for such a mud 
wall technique or a similar technique, rammed-earth architecture, is common 
throughout Africa (Guidoni 1978; and Wheaton personal observation 1966-1969), 
the presumed origin for some if not all of the inhabitants and probable 
builders of the structures at Curriboo and Yaughan. 



194 



If it can be assumed that the trench structures were mostly representative of 
mud wall construction, then the roofs were probably of thatch tied to roof 
supports, which were in turn attached to the tops of the posts protruding 
from the walls. Normally in this type of architecture the roof line extended 
well beyond the wall to protect it from rainfall and provide shade; this 
resulted in a drip line well beyond the wall. Soil disturbance by agricul- 
ture and stripping at 38BK75 and 38BK245 and root and other natural distur- 
bances at 38BK76 prevented identification of any drip lines. 

With such mud wall and thatch roof structures, one should also expect to find 
evidence of a rather high burn down rate even if wattle and daub chimneys 
were used. The absence of chimneys on the domestic structures or of any evi- 
dence that the structures burned down makes the probability of interior 
hearths remote, since one would expect at least one thatched roof to catch 
fire. Two obvious conclusions can be drawn in such a situation: some kind 
of exterior hearths were used, or all cooking was centralized and not located 
during fieldwork. Unfortunately, the evidence for permanent outside hearths 
was slight. Only one permanent hearth was found at Site 38BK75, located out- 
side of a posthole structure (Feature F25). At Site 38BK76, on the other 
hand, shallow depressions filled with soil and ash may have been temporary 
hearths (Features F33 and F82). Due to ground disturbance from cultivation, 
root and small animal action, and destruction from heavy equipment at 38BK245 
before mitigation, it was impossible to establish the presence of temporary 
hearths at that site without excavating outside the limits of the project 
area. Negative evidence against centralized cooking, at least at Site 38BK76 
where virtually all structures were exposed, tended to indicate that it was 
likely temporary exterior hearths, possibly only a few stones to set a pot 
on, were the primary source of cooking fires at the slave quarters. Along 
these same lines, it is interesting to note that modern pottery firing of the 
Yoruba in Nigeria involves simply piling pots and lightweight fuel together 
on the surface of the ground. Perhaps all fire locations at the slave quar- 
ters for cooking and pottery firing were temporary in nature. 

In order to test the significance of this apparent difference in mean dis- 
tances, an _F test was conducted. Although an F_ test is normally used to test 
variance it can also be used to compare means (Downey and Heath 1974:211). 
The F score was then converted to a t score using the formula _t ■ F_ (Downey 
and Tfeath 1974:215). The formula used was the following: 



mean-square between groups 
F_ score = 

mean-square within groups 



where: 

sum of squares 



mean-square = 

degrees of freedom 



195 



, 2 (C« z <C*T> 2 

between sum of squares = ?xk = £ 

\ n N 



2 2 2 

within sum of squares = ^x w = ^x t - ^ x b 

<- 2 < » ^ x)2 

total sum of squares = > x^ = ^X' 1 

M 

The data developed utilizing this method for all trench and posthole 
structures at all sites is presented in Table 20. 

TABLE 20. Postmold Distance Data 
All Trench and Posthole Structures 

Trenches Postholes 

Sum = 259.80 = 322.80 

Mean Distance = 2.2205 = 4.2474 

Sum-of-Squares = 615.20 = 1478.92 

Number of cases ■ 117 = 76 

Between sum-of-squares ■ 189.2725 

Within sum-of-squares = 146.1804 

Total sum-of-squares = 335.4529 





df 


sum-of-squares 


mean-square 


Between groups 
Within groups 


1 

191 


189.2725 
146.1804 


189.2725 
.7653 


F = 189.2725 = 247.3044 
.7653 









t = F = 15.7259 

t ■ 12.706 at 1 degree of freedom at .05 level of probability 



196 



With this data it can be stated that the apparent difference in mean dis- 
tances is probably a true difference. Whether or not this was a consequence 
of different superstructures, or whether it reflects adaptation over time to 
the difficulty of digging long trenches into the hard red clay subsoil, can- 
not be decided by statistics. However, based on the other data presented 
above, it is felt that mudwalled and frame structures are the best 
explanation. 

Time, at least superficially, correlates with trench and posthole structures 
and, therefore, their post distances. Site 38BK75, the most recent site, had 
three posthole structures and only one trench structure. Site 38BK76 had 
nine trench structures and five posthole structures. The stripped area at 
38BK245, the earliest site, had five trench structures and one or possibly 
two posthole structures. Further corroboration of the relative ages of the 
two kinds of structures and, hence, posthole distance was provided by the 
placement of posthole structures on top of trench Structures 75B?, 76Bi, and 
76Dx- 

In order to check whether there is a difference between post distances within 
trench and posthole structures from site to site which would imply idiosyncra- 
tic differences probably not related to superstructure construction, compari- 
sons were tested between trenches at 38BK76 and 38BK245 and between postholes 
at 38BK75, 38BK76, and 38BK245. 

The mean distances between posts in trenches at 38BK76 and 38BK245 were 
2.1327 feet and 2.2889 feet, respectively. This difference is small and is 
not statistically significant. The mean distance between posts in postholes 
at 38BK75, 38BK76, and 38BK245 are 3.9323 feet, 4.1974 feet, and 6.2000 feet, 
respectively. Here there is obviously a greater variability than for 
trenches. These differences probably are not statistically significant due 
to small sample size, but since the distance at 38BK245 was almost double 
that at 38BK75, something does seem to be occurring at the posthole struc- 
tures which does not appear in the trench structures. This variability may 
be due to the availability of building materials (e.g. longer beams which 
could span larger distances, available to the builders at 388K245, perhaps), 
or to differences in the function and, therefore, construction of posthole 
structures. A wider spacing might be presumed for open-sided thatch roofed 
sheds or work areas as opposed to enclosed frame cabins. In any case, there 
was roughly three times as much variability in post distance in posthole 
structures than in trench structures as illustrated by their variances, 
1.4383 for posthole structures and .3303 for trenches. 

Orientation of the structures may also provide a clue to differences in func- 
tion and date of construction within the sites. There were obviously dif- 
ferences between the orientation of structures such as 76L and 76J which were 
different from the other structures at 38BK76, or the structures at 38BK245, 
which were different from those at 38BK75 and 38BK76. Such differences are 
probably due to topographic conditions or the owner's personal tastes. It is 
unfortunate that so few of these structures were hand excavated to provide 
data on function. Beneath this readily identifiable difference in orienta- 
tion was a deeper pattern of orientation, which permitted certain structures 
to be offset at approximately 90°, but apparently within certain allowable 



197 



limits of variation. This seemed to hold true at sites 38BK76 and 38BK245. 
This offset might be likened to the buildings around a plaza where half the 
buildings are offset within a few degrees of 90° from each other, but the 
overall plan prohibits structures at 25°, 30°, or 45°. This deeper or more 
basic plan at the two sites can be examined by adding 90° to the lesser 
angles in order to "normalize" them for comparative analysis of the limits of 
allowable variation in the underlying plan. The converted structures were 
76J, 76L and all structures except 245C at 38BK245. Since structures at 
Sites 38BK76 and 38BK245 were either approximately parallel or approximately 
perpendicular to each other, the intent was to show that the same underlying 
plan was evident at each site. In order to compare the underlying plans 
within the sites, the data for Figure 70 was calculated. 

Figure 70 shows a graph of all structures; those which were offset 90° from 
the other structures have been converted to similar readings by adding 90°. 
As the figure illustrates, orientation falls into distinct groupings, 38BK75 
being the tightest at 128° and 38BK76 being the most variable. Indeed, 
38BK76 has two groups of orientation, one ranging from +U9°-+123° and a 
second from +134° to +145°. Site 38BK245 ranges from +143° to +147°. 

T- tests were run on the data to determine if the observed differences were 
statistically significant. The structures oriented from 134° to 145° at Site 
38BK76 were compared to those at 38BK245 and not found to be significantly 
different from each other. However, when the structures oriented from 119° 
to 123° were compared with the others at 38BK245, the result was a t-score of 
25.3872 with one degree of freedom. This is significant at the .001 level. 
It is not surprising, therefore, that the two groups at 38BK76 are also sig- 
nificantly different from each other at the .001 level. The structures 
oriented from 119° to 123° at 38BK76, however, were not statistically differ- 
ent from those at 38BK75, which all measured 128°, although this may be the 
result of small sample size in the later group. 

The differences noted in the structures at 38BK76 were formally compared to 
determine whether there may have been two groups or building episodes at the 
site. Structures 76C, 76D, 76E, 76F, and 76M (Figure 26) on the west end of 
the site were included in Group A. Structures 76A, 76B, 76G, 761, 76J, 76K, 
and 76L on the east end of the site were included in Group B. The mean 
orientation of Group A was 121.5° and that of Group B was 139.6°. This 
difference proved to be significant at the .001 level, meaning that the 
average orientation of the buildings probably represented two different 
populations of building orientation. This can be interpreted to mean that 
the buildings were probably built by two different groups of people at two 
different times. Mean ceramic dates developed from excavated and surface 
materials in the vicinity of the structures are presented in Table 21. 



198 



150°- 



140 



o .. 






5 130° 

e 
.£ 

w 

O 



120 



o ., 



IS 



o 

<D 

0J 



CO 



o 

CD 



o 

CD 
C4 



O 



o 
<o 

r- 

o 
in 



10* 



<o 



at 



m i*- 



10 



^ 



— v 

GROUP A 



.A. 



GROUP B 



FIGURE 70 

House Orientation at 

Yaughan and Curriboo 

Plantations 



199 







TABLE 


21 








38BK76 Structure 


Groups 


A and B 






Mean 


Cerami 


c Dates 




Group 


A 






Group 


B 


76C 


1770.6 






76A 


1773.4 


76D 


1770.6 






76B 


1787.6 


76M 


1770.6 






76 G 


1793.8 


76E 


1765.1 






76K 


1769.3 


76F 


1731.0 






76 L 
761 
76J 


1769.3 
1793.8 
1767.3 



Mean 1761.6 1775.7 

Range 1731.0 - 1770.6 Range 1767.3 - 1793.8 



Although the dates in Table 21 are often based on very few sherds or on 
surface material which may have been moved within the area covered by a 
group, there is clearly a trend for Group B to be later than Group A. These 
data on orientation would indicate that two building episodes were repres- 
ented at the site and that Group A was built before Group B. 

Data were examined to determine if post distance increased with time as has 
been hypothesized. The mean distance of Group A was 3.41 feet and for Group 
B it was 2.70. Although a difference in distances was noted, it was not 
significant. Moreover, the means ran counter to the hypothesis that post 
distance increased over time, and it was felt that something other than time 
was affecting the post distances at 38BK76. This difference may reflect a 
difference in function. The eastern and later structures were, for the most 
part, slave cabins, and the western and earlier structures were an overseer's 
cabin (perhaps the first "main" house) and specialized storage or activity 
structures requiring different architecture. 

Inspection of the floor plans of Structures 76D and 76M showed them to be 
unlike any other structures at 38BK76. Structure 76D was built in two parts. 
The first part was a trench structure with a central line of postholes in- 
dicating floor supports or a poorly aligned interior wall, and the second was 
a pesthole addition continuing the structure to the west. Assuming, as has 
been assumed for all trench structures, that entrance to trench structures 
was gained through the narrow ends of the structures where there was no 
trench, then the floor plan of 76D was that of a shotgun house. Structure 
76M was aligned with Structure 76D, and the main portion of Structure 76M was 
the same width as Structure 76D. The houses were closely enough spaced so 
that they could conceivably have been connected. In any case, the size, 
orientation, and proximity of structures 760 and 76M indicated that they were 
probably associated, which resulted in a rather tight complex of structures. 
One of the other structures in Group A was also very different in floor plan 
from the other structures at 38BK76. This was Structure 76E, which had a 



200 



crude dividing wall trench across the center, a series of postholes parallel 
to the central trench, and a partial end trench. The floor plan resembled a 
crude copy of the early two bay structures at 38BK245. 

The artifacts in probable trash features associated with Group A structures 
also varied from those in the remainder of the site (Table 22). Specifical- 
ly, those features included Fl, F2, F4, F5, F7, F8, F10, and Fll. Trash fea- 
tures associated with Group B structures were F12, F13, F14, F33, and F82. 
The sheer number of trash pits in Group A may have indicated that it was 
inhabited earlier and longer than Group B. Using South 's (1977a) artifact 
pattern model, the Group A and Group B features are shown in Table 22: 

TABLE 22 

Site 38BK76 Structure Groups A and B 

Artifact Patterns of Associated Features 

Group A Group B 
# X # X 

Kitchen 

Architecture 

Furniture 

Arms 

CI othi ng 

Personal 

Tobacco Pipes 

Activities 

1555 Z9T 



Three things should be noted from this data. First, the amount of artifacts 
was over five and one half times larger in Group A than Group B, although the 
number of features was less than two times higher. Second, the percentage of 
kitchen artifacts was higher and the percentage of architecture artifacts was 
lower in Group B than Group A. Third, only Group A had furniture and cloth- 
ing artifacts, although their percentages were very low. 

More importantly, the Kitchen Group artifacts (Table 23) were separated into 
colonoware, nonlocal ceramics, wine bottle glass, and other kitchen arti- 
facts. No tableware was recovered from the features in either group and only 
one iron kettle fragment was found in F8, Group A. The remainder of the ar- 
tifacts consisted of various pieces of non-olive green bottle glass. 



1350 


81.08 


260 


89.35 


258 


15.50 


24 


8.25 


2 


.12 





- 





- 





- 


8 


.48 





- 





- 


1 


.34 


44 


2.64 


5 


1.72 


3 


.18 


1 


.34 



201 



TABLE 23 

SUe 38BK76 Structure Groups A and B 

Kitchen Artifacts of Associated Features 



Group A Group B 

# % # 



1096 


81.25 


242 


93.08 


39 


2.89 


9 


3.46 


89 


6.60 


8 


3.08 


126 


9.34 


1 


.38 



Colono 

Non-local Ceramics 
Wine Bottles 
Other 

Total r3HD 2W 



Table 23 indicates that there was relatively more Colono in the Group B fea- 
tures than in the Group A features, nonlocal ceramics were about the same, 
wine bottles were twice as frequent in Group A, and Group A had an overwhelm- 
ing amount of other bottle glass. A chi -square test was run comparing these 
frequencies and the overall pattern was significantly different at the .001 
level . 

Since the greatest percentage differences were for olive green glass and 
other kitchen artifacts, mostly other bottle glass, it was decided to lump 
these two categories together to compare them with 38BK75, Spiers Landing and 
South 's (1977a) Carolina Artifact Pattern (CAP) in hopes of showing their af- 
finity to an Anglo-American pattern. Since South 's artifact patterns include 
very little Colono, it was also necessary to collapse the Colono and nonlocal 
ceramics categories in order to make a comparison. For comparative purposes, 
only trash features were used at 38BK75 and 38BK76 and Spiers Landing. The 
Revised CAP data included all material from Brunswick S7, S10, and S25. The 
reasons for exclusion of the other sites in South' s original pattern are 
given elsewhere in this report. The results are presented in Table 24. 

TABLE 24 
38BK76 Structure Groups and Various Sites 
Ceramics and Glassware of Associated Features 

38BK76 38BK76 38BK75 
Group A Group B 75B 
# % # % § % 
All Ceramics 1135 34.14 251 96.54 936 $4.35 

All Glassware 214 15.86 9 3.46 56 5.34 



202 



Revised 
CAP 

# % 



Spiers 
Landing 

# % 



All Ceramics 23670 71.81 

All Glassware 9290 28.19 



913 
119 



88.47 
11.53 



It should be noted that although all of the sites varied considerably from 

the CAP, the pattern at 38BK76 Group A was more similar to the CAP than that 

of any of the other sites. Chi -square tests were used to establish whether 
the differences were significant. 



Sites 



TABLE 25. Chi -square Values 
Comparing Ceramics and Glassware 

Chi -square Value 



Significant 
Difference 



38BK76 Group B vs. 38BK75 
38BK76 Group A vs. 38BK76 Group B 
38BK76 Group A vs. Spiers Landing 
38BK76 Group B vs. Spiers Landing 
38BK76 Group A vs. South 's 
Revised CAP 



1.9956 
28.0841 

9.1237 
15.1510 

98.2493 



none 
at .001 level 
at .05 level 
at .001 level 

at .001 level 



d.f. = 1 



p = .10 at 2.706 



p = .05 at 3.841 and .001 at 10.827 



The significance of the differences in relative frequencies of Kitchen Group 
artifacts was supported by the chi -square values. It was therefore concluded 
that 38BK76 Group A had significantly more glassware than 38BK76 Group B, 
38BK75, and Spiers Landing, and significantly less than the Revised CAP. 
While this did not establish that Group A and the CAP were similar, it did 
imply that Group A was somewhere between the slave occupations and the CAP. 
This can be explained if it is understood that glassware was essentially a 
luxury item and was curated as such during the eighteenth and early nine- 
teenth centuries. As a luxury item, glassware would not have been as easily 
available to slaves as it would have been for overseers and owners. If the 
structures in 38BK76 Group A were used primarily by an overseer or owner, 
then one would expect more glassware there than at 38BK76 Group B, 38BK75, or 
Spiers Landing. As glass became more readily available in the nineteenth cen- 
tury, one should expect to find relatively more at later slave sites than ear- 
lier sites. This may be the reason for the higher level of glass at Spiers 
Landing than at 38BK76 Group B and 38BK75. 



203 



The difference between 38BK76 Group A and the Revised CAP should not have 
been as great as it was, if the structures at 38BK76 Group A were inhabited 
solely by an overseer or owner. Based on the vast amount of Colono and 
documentary evidence, it is likely that 38BK76 Group A was not inhabited 
solely by an overseer or owner during its entire occupation. It can be 
hypothesized that only Structures 76D and 76M may have been inhabited by an 
overseer or owner at the same time as Structures 76C, 76E, and 76F were used 
by slaves and for other functions. Documentary evidence and mean ceramic 
dates point to a building date in the 1780s for the quarters at 38BK75 and 
the owner's house west of 38BK75. If this was indeed the case, it is likely 
that structures 76D and 76M may have been occupied by slaves from the 1780s 
onward. Both the fact that not all of the structures at 38BK76 Group A were 
ever inhabited by the overseer or owner and that structures 76D and 76M may 
have been occupied by slaves for the latter part of their existence would 
naturally obscure a CAP type of pattern. There was also a suggestion from 
documentary evidence that the first years of the plantation may have involved 
the part-time employment of an overseer at Yaughan. Such part-time occupa- 
tion could easily be partially masked by later slave occupation. 

Before continuing with a discussion of structure size and shape, the data on 
foundation type, posthole distance, and orientation should be summarized and 
related to the original hypotheses. It was evident from the general trend of 
foundation types that trench foundations were generally earlier than posthole 
foundations at 38BK75 and 38BK76. This was further supported by postholes 
intrusive into trench structures. Post distance was wider in posthole houses 
and this, coupled with other data, led to a conclusion that the superstruc- 
tures at the sites shifted from mudwall to frame in the eighteenth century. 
Evidence has also been presented that there were two building episodes at 
38BK76; this indicated that there was an increase 1n the size of the slave 
quarter around the time that Thomas Cordes began to live at the plantation in 
the early 1780s. The earlier portion of 38BK76 may also have housed the 
overseer in the early years of the plantation. At 38BK245, brick construc- 
tion superceded trench construction for special use structures, as shown by 
the naval stores trench structure and the administrative brick structure. 

Size and Form 

Size of the buildings was analyzed in two ways. In the first comparison, a 
basic construction unit was examined. This was the bay, or the largest 
single area encompassed by rows of posts whether in trenches or postholes. 
This is termed a "building unit" by Oeetz (1977:149-150). In the second 
comparison, whole structures, which might have included from one to three 
units and any porch or shed additions, were used. Both sets of comparisons 
proved useful in understanding the architecture. 

The bay or building unit comparison resulted in two statistically distinct 
categories; these were small building units and large building units (Figure 
71; Table 26) . The small building units averaged 145 square feet. South' s 
(1977a: 119) method for determining a range around the mean resulted in a 95 
percent probability range of 120.7 to 169.3 square feet for the smaller 
units. The larger units averaged 256 square feet with a 95 percent probabi- 
lity range of 185.8 to 326.2 square feet. 



204 



UJ 

o < 

K <T 



CO 

< 

oe 

< 

UJ 

CE 

t- 
U 

<r 
co 



< 
O 



ii 

Sir 
■* o 



K 
3 

o 

UJ 
UJ 



tr 



s 



■a en 

C 3 

fl ifl fl 

TO >» 

•i- s- -a 

c < c oo 

=> IB C 

i-H <u o 

r^ oi S. o 1- 

C 3 O -M 

UJ •!— 4-> J2 fO 

Q£ T3 U •«- +J 

=3 .— 3 S- C 

t3 -i- S_ 5- it) 

« 3 +-> 3 >— 

U_ CQ CO <_> Q_ 



r 




«n » m cm — 

1 ■ I ■ ■ 



^ UJ 



UJ 

cr 

< 

O 
CO 



< 

UJ 

cr 

o < 

o 



UJ 

3 



5 



UJ 

o 



m 
+1 



UJ 
(9 

^5 



o 

-o 

(M 



CO 

< 



cr 

< 



3 



CD 



5 



r 



_i < 

« 
at 






<M — 
1 I 



UJ 
UJ 



ac 

O 
CO 



-8 



cr 

< 



STRUCTURES 



o 
o 



STRUCTURES 



TABLE 26 
Widths, Lengths, and Areas of Building Units 



205 



WIDTH 



.ENGTH 



W/L 



AREA 



9.5 


13.0 


.73 


124.0 


10.25 


12.75 


.80 


130.7 


9.75 


14.5 


.67 


141.4 


10.75 


13.5 


.80 


145.1 


U.5 


13.0 


.83 


149.5 


11.5 


13.0 


.38 


149.5 


10.0 


15.0 


.67 


150.0 


9.75 


15.5 


.63 


151.1 


L1.0 


12.5 


.88 


153.0 


12.0 


13.0 


.92 


156.0 







FOUNDATION 


TYPE STRUCTURE 


LOCATION 


TYPE* 


mud wall hut 


76A 


T 


frame hut 


75B 2 


P 


mud wall hut 


76G 


T 


frame overseer's house 


76D 


T 


mud wall hut 


76E 


T 


frame overseer's house 


76M 


P 


mud wall hut 


245E 


T 


mud wall hut 


76L 


T 


frame hut 


75B, 


D 


frame hut or shed 


761 


P 



10.5 


13.6 


.79" 


145.0 








MEAN 




Mean 95', Range 


















3.5- 


11.1- 


.54- 


120.7- 












12.7 


16.1 


1.04 


169.3 












11.69 


18.13 


.64 


212.0 


mud wall hut 






76B, 


T 


12.0 


18.5 


.65 


222.4 


mud wall hut 






76F 


T 


11.5 


21.5 


.53 


247.25 


mud wal 1 hut 






76C 


T 


14.0 


18.75 


.75 


262.5 


mud wall hut 






76K 


T 


13.5 


20.0 


.68 


270.0 


mud wall hut 






245D 


T 


13.75 


19.75 


.70 


271.6 


mud wall hut 






245D 


T 


14.0 


19.5 


.72 


273.0 


frame or mud 


wall 


hut 


245B 


T 


14.0 


20.5 


.68 


287.0 


frame or mud 


wall 


hut 


245B 
MEAN 


T 


13.1 


19.5 


.57 


2bb.O 




Mean 95", Range 


















10.3- 


16.8- 


.50- 


185.8- 












15.9 


22.4 


.84 


326.2 












3.5 


10. 


.85 


35.0 


frame shed 






76J 


P 


10.0 


10.0 


1.00 


100.0 


frame shed 






75C 


P 


10.0 


16.5 


.61 


165.0 


office 






24 5C 


B. 


18.0 


18.0 


1.00 


324.0 


frame hut or 


shed 




76B 2 
245H 


P 


23.75 


41.5 


.57 


986.6 


frame barn 






T 


8.5 


18.0 


.47 


153.0 


frame or open 


shed 


245G 


P 


10.5 


17.0 


.62 


178.5 


frame overseer's house 


760 


P 



= trench, P - posthole, 3 = brick 



206 



It was noticed that the large house units seemed to occur with trench founda- 
tions, whereas the small units occurred with both foundation types. A Fisher 
exact test showed that the probability of such a pattern occurring by chance 
was only .0686, or very close to the .05 level of significance. It was there- 
fore concluded that small units were used when constructing trench and post- 
hole structures and larger units were used only for trench structures. Al- 
though neither the building units nor foundation type were one to one indica- 
tors of chronology, there were more small posthole structures at 38BK75 than 
at the earlier sites. 

The building units could be and were added together, end to end, to produce 
larger multiple bay structures. Such structures resembled the typical shot- 
gun pattern (as noted for Structure 76D). It is interesting that with all of 
the possible combinations of trench, post, and large and small units, only a 
few patterns for multiple structures were actually employed. These total 
structures could be statistically grouped into only three sizes: small struc- 
tures which averaged 145 square feet; medium-sized structures which averaged 
257 square feet; and large structures which averaged 552 square feet (Table 
26). That these total sizes were not simply multiples of the unit sizes was 
explained by inclusion of porches and other additions to the core structure. 
At the 95 percent level of probability none of the total structure floor 
space ranges overlapped (Table 26). 

The literature, although incomplete, was consulted to discover whether the 
units fit patterns on other slave, Anglo-American, or African sites. The 
widths, lengths, width to length ratios, areas, and house types and locations 
from various sources (Deetz 1977; Fairbanks 1972; Vlach 1977; Otto 1975; 
Genovese 1976; Glassie 1975; Mull ins 1980; Baker 1978; Drucker and Anthony 
1979, Leavitt 1980, Kelso 1980) are presented in Table 27. Some of this 
information was from single structures and some involved theoretical con- 
structs or ideal building units developed by the various investigators. 

Comparing areas produced some exact or nearly exact matches between the 
Cooper River structures and the comparative literature. Table 27 presents 
four occurrences of 144 square feet; these were a shotgun house in New 
Orleans, two freed slave houses in New England, and a "typical" West African 
house. An area of 144 square feet was only one square foot less than the 
small bay average. Two occurrences similar to our large bay structure were 
found; these were a "typical" African structure in the Caribbean and Deetz' s 
and Glassie' s "typical" Anglo-American building unit, at 252 and 256 square 
feet, respectively. The 252 square foot unit was only 4 square feet from our 
average, and 256 was, of course, exactly the same. 

These comparisons were only suggestive, especially because the size of struc- 
tures was easily limited by the amount and kind of materials available regard- 
less of the cultural biases of the builder. Therefore, size may not be the 
best criteria for indicating the mind set which conceived the structures. 
Comparison of the length to width (W/L) ratios, which should remain constant 
whether a building unit is large or small, may be a more lucrative approach 
to examine cultural differences and similarities. 



TABLE 27 
Various Building Unit Dimensions Compared 



207 



W/L 
WIDTH LENGTH RATIO AREA 

9.0 12.0 .75 108.0 



3.2 
12.0 
12.0 
12.0 
12.0 
12.0 
12.0 
12.6 
12.6 



14.0 



16.0 



16.0 



16.1 
12.0 
12.0 
12.0 
12.0 
14.0 
14.0 
16.1 
16.4 



11.3 21.3 



18.0 



.51 

1.00* 

1.00 x 

1.00 x 

1.00 x 

.86 

.86 

.78 

.77 



16.0 1.00 



132.0 

144.0 s 

144.0 s 

144.0 s 

144.0 s 

168.0 

168.0 

203.0 

206.6 



.55 251.34 



.78 x 252. 1 - 



256. L 



18.0 



.89 288.0 



HOUSE TYPE 
LOCATION 

freed slave - 
Parting Mays 

slave - Kingsley 
Plantation 

freed slave - 
New Orleans 

freed slave - 
Parting Ways 

"typical" - 
West Africa 

freed slave - 
Black Lucy's Garden 

shotgun - 
Port au Prince 

slave - 

Altama Plantation 

slave - Kingsley 
Plantation 

slave - 
Littletown Quarter 



domestic slave - 
Sinclair 

"typical" African 
Caribbean 

"typical" Anglo- 
American Unit 



slave "ideal" 



GENERAL 
CONSTRUCTION 
OATE 
(CENTURY) 


REFERENCE 


18th 


Deetz 
1977:149-150 


19th 


Fairbanks 
1974:108 


19th 


Vlach 
1977:52 


18th 


Deetz 
1977:150 




Deetz 
1977:150 


19th 


Baker 
1978:8 


19th 


Vlach 
1977:52 


19th 


Otto 
1975:104 


19th 


Fairbanks 
1974:108 


18th 


Kelso 
1980 per. 
communica. 


19th 


Mu 11 1 ns 
1980 




Genovese 

1974:528 


19th 


Deetz/ 
Kniffen 
1977:149-150 
1965:565 


19th 


Otto 
1975:103 



TABLE 27 (continued) 



208 



15.5 


18.9 


.32 


292.6 


slave - 
Spiers Landing 


19th 


Orucker S 

Anthony 

1979:91 


13.1 


23.0 


.57 


301.3 


slave - 

Hampton Plantation 


19th 


Mull ins 
1980 


16.0 


20.0 


.80 


320.0 


slave "ideal" 


19th 


Otto 
1975:103 


17.0 


20.0 


.85 


340.0 


slave - 
Cannon's Point 


19th 


Otto 
1975:112 


16.4 


22.2 


.74 


364.1 


slave - 

Littletown Quarter 


18th 


Kelso 
1980 per. 
communica. 


20.0 


20.0 


1.00 


400.0 


slave - 

Shirley Plantation 


19th 


Leavitt 
1980 per. 

communica . 


20.0 


22.5 


.89 


450.0 


slave - 
Cannon's Point 


19th 


Otto 
1975:111 


14.8 


16.4 


.90 


242.7 


overseer 
Cannon's Point 


19th 


Otto 
1975:118 



"Indicates structural dimensions used to compare width /length ratios over time 
* Within Cooper range 
3 Smal 1 bay 
L Large bay 



209 



Table 28 summarizes the W/L ratios already presented in Table 26 (the Yaughan 
and Curriboo structures) and in Table 27 (structures from the comparative 
literature). The average ratios from Yaughan and Curriboo included only what 
were considered to be slave dwellings and did not include sheds, barns, the 
office at 38BK245, or the hypothesized overseer's house at 38BK76 (Structures 
76D and 76M). The average ratios from the comparative literature did not 
include the "ideal" slave dwellings, the "typical" West African house, or the 
"typical" African-Caribbean house, since examples of actual slave occupied 
structures were comparatively numerous, and it was difficult, if not impos- 
sible, to determine how these "ideal" and "typical" ratios were arrived at. 
The "typical" Anglo-American building unit, on the other hand, was included 
in the third column because few such structures were included on the list and 
the basis for the unit has been thoroughly researched by Glassie (1975) and 
accepted by Deetz (1977). 



TABLE 28. Average Width/Length Ratios 
Slave 



Freed Slave 



Anglo-American 



Comparative 
Literature 


.76 




38BK76 Group A 


.75 




38BK76 Group B 


.68 




38BK75 


.84 




38BK245 


.69 




2 Standard 
Deviation Range 


.52- 


.92 


Observed Range 


.53- 


.88 


Total Curriboo 
and Yaughan 


.72 





.96 



1.00 



The first striking characteristic notable in Table 28 was that slave struc- 
tures were much more rectangular than freed slave and white structures, and 
that the freed slaves and Anglo-American ratios were not included within two 
standard deviations of the slave mean. Considering that a ratio of 1.00 is 
square, the freed slave and Anglo-American structures were nearly square, 
whereas slave structures definitely were not. It was also apparent that 
freed slaves, who presumably would have been much more acculturated than 
slaves, built structures of essentially the same shape as their white neigh- 
bors. If freed slaves were indeed more acculturated and this was reflected 
in the width/length ratios of their houses, then such a scale might be useful 
in examining the extent of acculturation evident in structures built and oc- 
cupied by slaves. 



210 



Table 28 shows that the different portions of the slave quarters at Yaughan 
and Curriboo seriated from the least square, 38BK76 Group B and 38BK245 to 
38BK Group A, to the most square (or approaching squareness) at 38BK75. This 
seriation obviously did not represent time, i.e., from the earliest to the 
latest, since a strictly temporal seriation based on MCDs was 38BK245 
(1760.5) to 38BK76 Group A (1761.5) to 38BK Group B (1779.2) and 38BK75 
(1789.8). 

Since the sites were not seriated by time and squareness seemed to be a 
measure of acculturation, it was concluded that the sites were seriated by 
the degree of acculturation (Table 28) and that 38BK76 Group A again appeared 
to be more acculturated than 38BK76 Group B. 

Many more separate houses would have to be examined with better controls be- 
fore statistically meaningful comparisons could be made. But it does seem 
that one of the first avenues for investigating the origins of slave archi- 
tecture in the eighteenth century might be West Africa and the Caribbean as 
noted from this discussion of shape and in Table 26 and 27. 

Before leaving the subject of structures, another facet of plantation culture 
can be examined through study of the structures. Documentary evidence has 
shown that eighteenth century plantations in coastal South Carolina were not 
as specialized as they were to become by the nineteenth century. In particu- 
lar, early in the eighteenth century when plantations were first being set- 
tled, plantation owners experimented with various methods of producing in- 
come. Samuel Cordes at Curriboo plantation was no exception. 

The structure associated with naval stores production at Structure 245C was 
one example of such activity. The brick clamp at Structure 245K provided 
another example. Samuel Dubose (n.d.), in his book on St. Stephens' parish, 
noted that Samuel Cordes attempted to obtain the contract for the bricks used 
to build the parish church. However, he was unsuccessful because of the poor 
quality of his brick. The brick clamp at 245K may have represented an ar- 
chaeological manifestation of this effort. Since Cordes was unsuccessful, it 
could be assumed that he used the bricks himself. As a result of the severe 
damage to Curriboo by excavation of the borrow pit, it was impossible to de- 
termine whether there may have been other brick clamps at the site. With the 
limited data at hand (Appendix E), it was hypothesized that the brick piered 
structure at Structure 245C was built with bricks from the clamp at Structure 
245K. This was tested by measuring the bricks at both loci and statistically 
comparing them. Although the two sets exhibited variability, it was hoped 
that they would be statistically similar enough to conclude that the clamp 
could have produced the bricks at the structure. The means of height, width, 
and length were all within one standard deviation of each other and were 
therefore similar enough to have come from the same brick population 
(Appendix E). It was therefore concluded that the bricks at Structure 245C 
could have come from the kiln at 245K. 



211 



Summary 

This chapter has presented and briefly discussed the structures and features 
at Sites 38BK75, 38BK76, and 38BK245. These structures were then compared 
from intra-site, inter-site, and off-site perspectives in order to examine 
the hypotheses established for the structures. It was shown that: 

- Posthole structures superceded trench structures at 38BK75 and 38BK76. 

- Brick construction superceded trench construction for major outbuildings 
at 38BK245 

- Posthole distances increased over time, implying a change in super- 
structure, probably from mudwall to frame. 

- No chimneys or interior hearths were used at the sites. 

- Orientation of the structures indicated two building episodes at 38BK76. 

- The structures at 38BK245 were the most regular in size, construction 
quality, and layout. Site 38BK76 showed great variability in these 
factors. 

- Structures 76D and 76M at 38BK76 were sufficiently unique in size, shape, 
construction, and artifact assemblages to indicate that they may have been 
used by an overseer during a portion of their occupancy. 

- Length to width ratios indicated a change from rectangular structures 
similar to Afro-Caribbean patterns to squarer structures similar to 
Anglo-American house patterns. 

- The owner of Curriboo Plantation engaged in the production of naval stores 
and bricks. 



212 



VIII. ARTIFACT ANALYSIS 

Introduction 

One of the goals of this project was to delineate and compare artifact pat- 
terns. To accomplish this goal and to make the product of our research com- 
parable to that produced by other researchers, it is essential to give expli- 
cit definition to the terms and artifact types discussed in this report. 
Existing published typologies have been used to the extent possible. In a 
few cases, typologies were developed, either because existing descriptions 
did not cover all of our material, or because the quantity of our material 
allowed refinements to be made in existing typologies. The artifacts are 
summarized by site and structure in Table 29 and Table 30. 

Table 29 presents the total artifacts from each site by artifact type. Each 
type is listed with its artifact group as established by South 
(1977a: 92-102). The three left hand columns on the table represent actual 
counts and the three right hand columns represent the relative amounts of 
each artifact type based on South 's (1977a) patterns. For this reason non- 
local ceramics which postdate the main occupation and unidentified iron were 
excluded from the artifact totals. Other changes from South' s (1977a) pat- 
terns are explained in detail in the following chapter. Table 30 represents 
the same kind of data following the same organization, except that selected 
structures are shown rather than complete sites. Structures 76A, 76B, 75B, 
and 245C were block excavations and 76D is included because it plays a promi- 
nent role in the discussion of site function. The totals for Structure 76D 
include all material excavated from features and surface collections in the 
vicinity of the structure. 

The main published sources used in the course of analyzing the 35,297 arti- 
facts retrieved were South (1977a) and Noel Hume (1978). These were supple- 
mented with more restricted analyses of particular categories, and a brief 
list of the most useful references in selected categories follows: 

Ceramics - Lofstrom (1976), Bartovics (1977), Miller and Stone (1970), 
Palmer (1976), Shepard (1965), Ferguson (1977), Gartley 
(1979), Handler (1963 and 1964), Handler and Lange (1978), 
Quimby (1973), Anthony (1979), Drucker and Anthony (1979). 

Gun Parts - Stone (1974), Ferguson (in South 1977b) 

Bottle Glass- Jones (1971), Douglas and Frank (1972) 

Architecture- Vlach (1977), Bonner (1945), Glassie and Kniffen (1972), 
Drucker and Anthony (1979) 

Furniture - 01 sen (1963) 

Tobacco Pipes- Petersen (1963), Walker (1967) 



213 



TABLE 29. Total Artifacts by Site and Type 







Site Total 


s 


Site 


Percentages 


ARTIFACT CATEGORY 


38BK75 


38BK76 


386K245 


38BK75 


38BK76 


38BK245 


Olive wine bottles 


400 


1631 


642 


6.37 


7.31 


11.02 


Bluish olive wine bottles 


15 


9 


5 


.24 


.04 


.09 


Other olive glass 


10 


40 


7 


.16 


.18 


.12 


Clear bottle glass 


68 


164 


23 


1.08 


.73 


.39 


Green tinted bottle glass 


63 


118 


12 


1.00 


.53 


.21 


Amethyst bottle glass 


1 






.02 






Table knives 


1 


2 


5 


.02 


.01 


.09 


Forks 


2 


2 


1 


.03 


.01 


.02 


Other tableware 


4 


3 


1 


.06 


.01 


.02 


Other kitchenware 


5 


5 


3 


.08 


.02 


.05 


Kettle fragments 


8 


14 


3 


.13 


.06 


.05 


Clothing iron 




1 






.01 




Non-local ceramics 


1022 


1627 


445 


16.28 


7.29 


7.64 


Colono pottery sherds 


2545 


15043 


3316 


40.55 


67.38 


56.91 


Catawba 


295 


141 


17 


4.70 


.63 


.29 


Non-local ceramics 














(unidentifiable) 


C 33] 


[ 42] 


C 3] 









KITCHEN GROUP TOTAL 



4439 



18800 



4480 



70.73 



84.20 



76.88 



Flat glass 
Cut nails 
Wrought nails 
Door locks 
Other architectural 



72 

37 

8 

3 



101 

175 

89 



114 

4 

80 

3 



1.15 
.59 
.13 
.05 



FURNITURE GROUP TOTAL 



.45 
.78 
.40 



1.96 
.07 

1.37 
.05 



hardware/objects 
Unidentified nails 


5 
1444 


10 
2265 


4 
760 


.08 
23.01 


.04 
10.14 


.07 
13.08 


ARCHITECTURE GROUP TOTAL 


1569 


2640 


965 


25.00 


11.82 


16.56 


Furniture hardware 


5 


12 


4 


.08 


.05 


.07 



12 



.08 



.05 



.07 



Musket balls and shot 
Gunflints and spalls 
Gun parts 



: 10 


2 


8 


.16 


.01 


.14 


l 


2 


8 


.02 


.01 


.14 




1 


1 




.01 


.02 



ARMS GROUP TOTAL 



11 



17 



.18 



.02 



.29 



[ ] not used in artifact pattern percentages 



TABLE 29. (continued) 



214 







Site Total 


s 


Site 


Percentages 


ARTIFACT CATEGORY 


38BK75 


38BK76 


38BK245 


38BK75 


38BK76 


38BK245 


Clothing buckles 


2 


6 


2 


.03 


.03 


.03 


Sewing equipment 














(thimble and pins) 




2 


1 




.01 


.02 


Silver plate buttons 


3 


3 


2 


.05 


.01 


.03 


White metal buttons 


8 


7 


5 


.13 


.03 


.09 


Lead buttons 


1 


10 


2 


.02 


.04 


.03 


Iron buttons 


5 


2 




.08 


.01 




Gold button 




1 






.01 




Bale seals 






2 






.03 


Glass beads 


1 


20 


3 


.02 


.09 


.05 


Brass and copper buttons 


12 


15 


4 


.19 


.07 


.07 


CLOTHING GROUP TOTAL 


32 


66 


21 


.51 


.30 


.36 


Coins 


1 






.02 






Keys 




3 


2 




.01 


.03 


Other personal items 


3 


2 




.05 


.01 




PERSONAL GROUP TOTAL 


4 


6 


2 


.06 


.03 


.03 


Pipe parts 


182 


744 


306 


2.90 


3.33 


5.25 


Colono pipes 




8 


4 




.04 


.07 


TOBACCO GROUP TOTAL 


182 


752 


310 


2.90 


3.37 


5.32 


Clasp knives 


8 


13 


10 


.13 


.06 


.17 


Other tools 


3 


2 


4 


.05 


.01 


.07 


Colono toys 


1 


4 




.02 


.02 




Fishing gear 




5 






.02 




Unidentified - iron 


[380] 


[ 297] 


[112] 








Harness parts 


2 


3 


4 


.03 


.01 


.07 


Hoes 


7 


3 


4 


.11 


.01 


.07 


Other artifacts and 














Colono objects 


4 


6 


4 


.06 


.03 


.07 


Unidentified - lead 


8 


7 


1 


.13 


.03 


.02 


Unidentified - brass/copper 


1 


3 




.02 


.01 




ACTIVITIES GROUP TOTAL 


34 


46 


28 


.54 


.21 


.48 





TOTALS for South' s (1977a) 
Pattern 



6276 



22327 



5827 = 34430 



GRAND TOTAL (including 
metal and ceramics) 



6689 



22666 



5942 = 35297 



TABLE 30. Total Artifacts for Selected Structures 



215 



Structure 

m 



Total s 
Z*5C~ 



ARTIFACT CATEGORY 



Olive wine bottles 

Bluish olive wine bottles 

Other ol ive glass 

Clear bottle glass 

Green tinted bottle glass 

Table knives 

Forks 

Other tableware 

Other kitchenware 

Kettle fragments 

Nonlocal ceramics 

Colono pottery sherds 

Catawba 



1W 



75B 



260 


639 


266 




3 


11 


9 


5 


7 


22 


105 


55 


16 


78 


60 


2 




1 




1 


2 


1 




8 


3 






4 


3 


3 


173 


691 


526 


4586 


6762 


2174 


32 


85 


280 



22 

1 

1 

1 

1 

9 
25 



76D 



21 

1 
2 

1 



1 

18 

204 

3 



KITCHEN GROUP TOTAL 



5108 



8372 



3393 



60 



252 



Flat glass 
Cut nails 
Wrought nails 
Door locks 
Other architectural 



34 
22 

7 



31 

132 

53 



68 

35 

6 



52 

19 
3 



hardware/objects 
Unidentified nails 


1 
865 


3 
902 


4 
1290 


2 

132 


2 

68 


ARCHITECTURE GROUP TOTAL 


929 


1121 


1403 


208 


71 


Furniture hardware 


2 


7 


3 






FURNITURE GROUP TOTAL 


2 


7 


3 






Musket balls and shot 
Gunflints 


1 


1 
1 


8 


2 




ARMS GROUP TOTAL 


1 


2 


8 


2 





216 



TABLE 30. 


(continued 


) 










Structure 


Totals 




ARTIFACT CATEGORY 


76A 


76B 


75B 


246C 


76D 


Clothing buckles 

Sewing equipment 

Silver plate buttons 

White metal buttons 

Lead buttons 

Iron buttons 

Gold button 

Glass beads 

Other clothing articles 

Brass and copper buttons 


2 

1 

1 

1 
2 

1 


3 

3 
3 
5 
2 

7 

10 


2 

3 

7 
1 
4 

1 
2 
9 


1 


1 


CLOTHING GROUP TOTAL 


8 


33 


29 


1 


2 


Coins 

Keys 

Other personal items 


2 


2 


1 
3 






PERSONAL GROUP TOTAL 


2 


2 


4 






Pipe parts 
Colono pipes 


201 
2 


295 
3 


123 


10 


12 


TOBACCO GROUP TOTAL 


203 


298 


123 


10 


12 


Clasp knives 

Other tools 

Colono toys 

Fishing gear 

Harness parts 

Hoes 

Other artifacts and 

Colono objects 
Unidentified - lead 
Unidentified - brass/copper 


6 
3 

5 
3 


6 
2 

2 
2 

4 
2 
2 


3 
3 

1 

2 

6 

4 
4 
1 


3 

1 
1 


1 


ACTIVITIES GROUP TOTAL 


18 


20 


24 


5 


1 



TOTALS (for South' s 
(1977a) pattern 



6271 



9855 



4987 



286 



339 



TABLE 30. (continued) 



217 



ARTIFACT CATEGORY 



76T 



Structure Percentages 
~76B 75B Z4~5C" 



T5D 



Olive wine bottles 
Bluish olive wine bottles 
Other olive glass 

bottle glass 

tinted bottle glass 

knives 



Clear 
Green 
Table 
Forks 
Other 
Other 



tabl eware 
kitchenware 
Kettle fragments 
Nonlocal ceramics 
Colono pottery sherds 
Catawba 



4.15 

.14 
.35 
.26 
.03 

.02 
.05 
.06 
2.76 
73.13 
.51 



6.48 
.03 
.05 

1.07 
.79 

.01 



.03 

7.01 

68.61 

.86 



5.33 
.06 
.10 
2.10 
1.56 
.02 
.04 
.16 

.06 

10.55 

43.59 

5.61 



7.69 

.35 

.35 

.35 

.35 

3.15 
8.74 



6.19 

.29 

.59 
.29 



.29 

.29 

5.31 

60.18 

.88 



KITCHEN GROUP TOTAL 



81.45 



84.95 68.04 



20.98 



74.34 



Flat glass 
Cut nails 
Wrought nails 
Door locks 
Other architectural 
hardware/objects 
Unidentified nails 



.55 
.35 

.11 



.02 
13.79 



.31 

1.34 

.54 



.03 
9.15 



1.36 
.70 
.12 



.08 
25.87 



17.13 

6.64 
1.05 

.70 
46.15 



.29 



.59 
20.06 



ARCHITECTURE GROUP TOTAL 


14.81 


11 


.37 


28 


.13 


72.73 


20.94 


Furniture hardware 


.03 




.07 




.06 




.29 


FURNITURE GROUP TOTAL 


.03 




.07 




.06 




.29 


Musket balls and shot 
Gunfl ints 


.02 




.01 

.01 




.16 


.70 




ARMS GROUP TOTAL 


.02 




.02 




.16 


.70 





TABLE 30. (continued) 



218 







Structui 


re Percentages 




ARTIFACT CATEGORY 


76A 


76B 


75B 


245C 


76D 


Clothing buckles 


.03 


.03 


.04 






Sewing equipment 


.02 






.35 




Silver plate buttons 




.03 


.06 






White metal buttons 




.03 


.14 






Lead buttons 


.02 


.05 


.02 




.29 


Iron buttons 




.02 


.08 






Gold button 


.02 










Glass beads 


.03 


.07 


.02 




.29 


Brass and copper buttons 


.02 


.10 


.18 






CLOTHING GROUP TOTAL 


.13 


.33 


.58 


.35 " 


.59 


Coins 






.02 






Keys 


.03 










Other personal items 




.02 


.06 






PERSONAL GROUP TOTAL 


.03 


.02 


.08 






Pipe parts 


3.21 


2.99 


2.47 


3.50 


3.54 


Colono pipes 


.03 


.03 








TOBACCO GROUP TOTAL 


3.24 


3.02 


2.47 


3.50 


3.54 


Clasp knives 


.11 


.07 


.06 


1.05 


.29 


Other tools 




.02 


.16 






Colono toys 


.05 




.02 






Fishing gear 




.02 








Harness parts 




.02 


.04 






Hoes 






.12 


.35 




Other artifacts and 












Colono objects 


.08 


.04 


.14 


.35 




Unidentified - lead 


.05 


.02 


.08 






Unidentified - brass/copper 




.02 


.01 






ACTIVITIES GROUP TOTAL 


.29 


.20 


.48 


1.75 


.29 



219 



As noted in the section of this report discussing the testing phase, no at- 
tempt will be made here to analyze the artifacts from the tree fall (38BK73). 
The following discussion only concerns material recovered from 38BK75, 
38BK76, and 38BK245. Further, it should be noted that only artifacts which 
contribute to the elucidation of our research goals or for which we have been 
able to add new insights are discussed here. All other historic artifacts 
are described in Appendix E. Prehistoric artifacts are presented in Appendix 
A. 

Ceramics 

Nonlocal (generally English) ceramics were represented by 3171 sherds from 
all three sites, or 9.0 percent of the total artifact assemblage for the proj- 
ect. Site 38BK75 had 1054 sherds, Site 388K76 had 1669, and Site 38BK245 had 
448. Of these sherds, 77 were unidentifiable as to type or ware because of a 
lack of glaze, burning, or because of their small size. These unidentified 
sherds are not included in the analyses which follow. 

A total of 63 nonlocal ceramic types were distinguished during the course of 
analysis and, along with a category of unidentified ceramics, were given com- 
puter code numbers for a total of 64 numbers. Three of these types were not 
considered to be pertinent to the main occupation of the sites. These were 
two plain whiteware types determined to date from the 1830s or 1850s to the 
present, and a third category including all other twentieth century ceramic 
types found on the surface and probably discarded as trash by the present 
inhabitants of the area. The remaining 60 distinguishable types differed oc- 
casionally from accepted usage, particularly when such divisions of estab- 
lished types proved to be helpful in understanding cultural processes. Table 
31 presents the ceramic types found at the sites and organized by ware. In- 
cluded on the table are mean ceramic dates (MCD) when these could be ob- 
tained, the reference for each MCD, and the total number of sherds at each 
site. 

The redwares and slipwares are discussed in some detail since it has been 
possible to identify types of these wares which seriate in time and may, 
therefore, be useful in comparative chronological studies. Redware is de- 
fined as those types which have a brick red, porous body and occasionally may 
have minor amounts of fine nonplastics. Surface treatment varied from un- 
slipped and unglazed types to types which had one or more slips and clear or 
opaque glazes. Some of these types are lumped together with buff bodied and 
variously decorated types by other investigators. However, some of the dec- 
orative techniques, motifs, and vessel forms were unique to the red bodied 
ceramics; therefore, these types are separately described here. 

The sorting criteria defined five types of redware, Thin Black Glazed, Thick 
Black Glazed, Clear Glazed, Trailed, and Funnelled Redware. Thin and Thick 
Black Glazed Redware were differentiated primarily on thickness and vessel 
form and may have represented varieties of the same type. The thin type 
would resemble Jackfield on complete cups and bowls when the paste was not 
visible. The thick type was used on larger vessels, mixing bowls, or deep 
plates, and was sometimes glazed only on one surface. Clear Glazed Redware 
had a clear probably lead glaze and resembled Thin Black Glazed in thickness 
and form. Trailed and Funnelled Redware may have been varieties of Clear 



220 



co 
cm 

CD 
CO 

co 



f o 
so CO 



so to 
co 



C) H -H ,-h 



p^ roffi 



«* -h 



CM 
en 



CO i-4 ^ <—< 



SO 

p>» 
:*£ 

CD 
CO 
CO 



CO CM CSJ CO 



SO SO CM -H CO 
vO CM 



CSJ rH CM rH CI i-H 



CM — I SO 

SO — i 



co 



-h o ^ 



in 

r». 
:*: 

CD 
CO 
CO 



~4 O 

—i CM 



SO CO 



o 
csj 



CM O 



SO —« 
SO 



o 



CO CSJ CSJ CO 


U1-HNOCA 


CO CM CSJ 


CO — H f-H 


«—* f— t CO H 


CM 


CSJ ^ 


CSJ 





*-< CO 



_< en co to 

Px ^- CM 



ir> — « 



OONON'tHrt 
CTi ^h «— » 

p»» 



T3 

c 

CL 

a. 
<. 

<v 

co 



o 

*E 

<1J 



0) 

u 






or 



o 

CM 

p»» 
p^. 
en 



4-> 

O 



CM 

P- 

p>» 



3 
O 
00 



c 

■— • O O *-H 

t-H t-< •<-> -H 

CM CM I/) CM 

p^ p*. "O r-. 

px px C P» 

en en *C Cn 

t— 4 «— t s CM •— • 

1- P^ 

JT -C 0) •• -C 

■«-> 4-> >— O ■»-> 

3 3 i— P-» 3 

O O i-W o 

co co s: -h <-o 



o 

C_3 



* * -K * 

o o o o 

CO CO CO P""» 

px p-» p«» px 



co co co co on 

CO to CO CO CO 

r»» r» p»» p"» r*. 



* * cn 
O O so 



o o o 

SO CO so 

p»» px px 



CO * * 

to o o 

px 



CM 



px 
cn 



3 
O 

co 



O CO 
SO cn 

Px Px 



CM 


«-H 


CM 


.— t 


—4 


*-4 


CM 


CM 


CM 


• • 


• • 


• • 


P- 


P». 


P»» 


P-» 


P— 


rx. 


cn 


Cn 


cn 



4-> 4-> -M 

3 3 3 

O O O 

CO CO CO 



co ^ p-~ co co co * 
cn p- so co co o o 
p>« p^ P». P»- px co 



o 

c 
o 



CO 



CO 



a. 



<X) 


















N 


















<a 


















^-» 


















CT> 


















I. 
















<u 


<U 
















£ 


<u > 
















<u o 


••-» o 
















3 »- 


•f- 












^^ 




f— £ 


j= a> 












-o 




CO U 


2 E 












^ 




>> 


O 












10 


•a 


JC »— 


C W 












2 


"O <u 


o o 


o x: 












<U tsl 


-M O. 


u 


»— 










O) N 


N <fl 


(U <T3 


Qi >-> <Q 




c 






•M 0J « 


<TJ r— 


3 i- -o 


3 i— 


-•-> 




5 






W» 4>> N r— 


•— cu 


.— u <u 


«— o 


c 




o 






<0 ^ "O «3 CO 


O) 


CO CO +J 


co a. 


a> 




L. 


E 




3 * 4I>- 


c *» 


c 




•*- 




CO 


ITS 




— WQJO'O 


ZS f— 


£ "O T- 


^— ^— 


4- 


C 




-C 




<o <u 


<o 


'_) <D "3 


<o <o o 


« 


£ 


en e 


.C CD "O i- (— 


C CO 


4-> CO Q. 


4-> -M 




<u 


M 


c 


<D 


«/) CJ (O ^ 


•*— 


<o <a "O 


c c 


e 


a. 


•^ 


•^ 


p— 


••- .* c a» i. 


(Q O) 


1.J3C 


4> <U 


t— 


o 


-M 


4-> 


to 


C u "~ i— ns 


.— ** 


o <u <o 


•r- "^ 


«3 


&. 


^-» 


•»-> 


L. 


<U <0 <*- C_3 S 


o. •»- 


CO Q I 


s- u 


p— 


3 


%. 


o 


3 


J= •— <v 


.c 




o o o_ 


LU 


CD Z 


CD 


cc ea cc 


2 





CO —■» 

0) CO 
Q. <U 

>> o- 

^^ 

co 

3 CO 

O 3 

f- o 
1- 1- 

> <o 



o <a 

'_ c 
CfflO 

<a c c 
a. a. 







01 










j* 










•^ 










* 










c 










o 










"O 










X •— 










O 2 <U 










,— O ^ 




01 


-o 


<D 


f— •— sz 




E 


r— 


s_ 


<U i— 3 




O 


0) 


« 


>- <u«— 


■o 


i. -a 


•t— 


» 


>- 


a> 


£ <u 


l+- 


a> 


•m e 


-a 


O i— -o 


J>^ 


•M 


£ ^ ai 


3 


>»-JO 0) 


u 


<o 


en *- <u 


O 


f- s_ en 


(O 


en 


•»- <0 !_ 


r— 


o « -o 


-^> 


< 


_ 1 Q O O Q. S UJ 






c 

<T3 



1) 

u 



o 

o. 



« 

2 

<u 

c 
o 

CO 



i. 

2 
c 
a> 



a> 



a> 



a> 

s 

E 

<u 

c 

o 



221 



LO 

cm 
59 

CO 



to 



Cn 



00 
CM 



CM 



CO 



CO 



LO 

en 

CO 



CO 



COW ififO 

ro en ~* 



en 

o 



co «*■ o <r> -h 



CM 



** ~* 



co cm 

LO 



lO 

co 

CO 

ro 



LOCMenr-^cn^f^eocn 

-* LO CM I — CM CM «3" ^ 

CM 



O 



o 



o^io^ooN^oai 


CO «a- O lo cm 


10 


«*• 


^ CO CM 


r*. hco - * ro c\i co •— i 


LO 


CVJ 


•— ( 


-H LO 


i— I ,— t CO 











<u 

C 



<u 
u 

c 

<D 
i- 
O) 
<4- 

<u 
on 



c_> 



CM 


CM 


— < 


»-* 


CM 


CM 


• • 


• • 


P>» 


f^ 


r*. 


P*. 


en 


en 



o 



o 
LO 



CM 


r- 


-H 


r-» 


CM 


en 


• • 


r-4 


P*. 




p»» 


</) 


en 


u 


fH 


•^ 




> 


J= 


o 


•»-> 


•4-> 


3 


s- 


O 


« 


LO 


CO 



p-» 
p» 
en 



(J 



> 

o 



CO 



<u 

c 
o 

4-> 
CO 



CO 

c 
o 

4-> 

co 



co 

c 
o 

4-> 

co 



CM ~* 

-H »— I 

CM CM 



c c 

co *r 

5- cm s- en 

<y .. (U 



c p«» p» 

<tj en en 

O •— 4 t— 4 

at 



i— O ■— Or-O-M-t-* 
^-P^f— P^i— P^ 3 3 

•»— en •»■" en •»— en o o 



c 
o 
u 



Q o * ifl w co ui in ifl « 
O 0000*-*OOOrf 

CO CO 00 CO CO CO CO CO 



« « « « * 
O CO to CO CO 

co co co co co 



* 
o 

LO 

p*. 



■K 

o 
to 



* 

LO 

p-. 



O CO 

LO CO 

p» l-» 



CO 

< 



co 

a. 
>) 



-a a» 








"O 


















a> *-> 








CO 


















•M C 4-> 








4-> 






4-> 


CO 










c •»- c w 








c 






C 


4-> 










1- <TJ •»- <TJ 








•^ 






•^ 


•^ 










« a i-i— 








<o 






i- 


J= 










Q.-o a> o_ 3 








ol co 






o. 


3 


CO 








■o e s c 








T3 E 










E 








C <Q O S- C 








c o 






t. 


■o 


o 


•o 






«s s- <u <: "o 






<TJ S_ 






CO 


c 


L. 


CO 




<u 


31 ■£=<«- 


a> 






Z J= 






<4- 


«o 


J= 


s_ 




u 


C "O CJ VI c 


("- 


c 


(Q 


(J 


■o 


c 


(/) 




a 


CO 


C 


c 


<U 25 4) >> C f- 


JO 


■f" 


-C 


CO >» 


<D 


•t— 


c 


CO 


>» 


•a 


t— 


a> 


3 O CH«— <*J <TJ 


s_ 


lO 


o 


3 r- 


cn 


ITS 


ITS 


3 




S 


«o 


•r* 


.— i. -o O J- i— 


<T3 


f"— 


o 


r— o 


■a 


r— 


&. 


r— 


"o 


o 




<TS 


CO CO Ul Q. r- o_ 


IE 


o. s 


CO Q. 


UJ 


o_ 


r- 


CO 


Q- 


Q. 


O. 


Ll. 



CO 



CO 

i. 

A3 



ITS 
CO 

o. 



co 



'co 



222 



Lf) 

CM 

CO 
00 

en 



lt> <*> en n 



o -h o en 

«*■ CVJ —l 



m *-* m 



CO 



r-. 

CO 
00 

m 



cm — * en csj »a- 

CVJ -4 



r~ *a- en vo in cvj oo 

«-H .—i «— • i>o ^ CVJ 



r-4 CVJ CVJ 



cr> 



m 

CO 
00 

en 



■n 



en 



in cvj 



cvj 
en 



in 
o 



o 



cvj cvj m in o 

CVJ CVJ ^ CVJ 



cvj oo r— o r-* cvj — ■* 
evi r-i cvj _* vo on — * 



^- on 



on 



co 

3 
C 



o 



en 



cu 
o 

c 
co 
s_ 
co 
<*- 
co 



CVJ 

en 



o 



Q 



en on on en en 
en on on on en 



cvj 



en 



o 



en on on en on en on 
on en on en on en en 



CVJ 

p«» 
en 



o 
o-> 



oo <^> 



co 
+-> 



E 

■V 

u 

c 



CO 

< 



co 

o. 



"O co 

CU N 
M <0 
•O i— 

.— C3 

.* co 

o <o <o "CJ ^» 

(O »— r- CO r— 
i— CO (J i— CO 

CO ■»- c 

^ 1- (TJ C 

11 U IS k 3 
C i- CO H- Li_ 



CO 



CO 
N 

CO .— 
N CD 

I— V. 

CS <TJ 
CO 

i. p- 

<o o 

CO 

»— "O 
O CO 

JO 

c S 
v o 
«J C_) 



CO 
N 

<o 

.— CO 
O tsl 

s_ ,— 
tO CD 
CO 

i— "O 

O CO 

4-> 

■o c 

CO t- 



1- 1- 



CO 
N 

<o 

3 

"O 

CO 

CO +■> 

CO N C 

N <TS •»- 

« r- I— 

CS "O 

"O CO 

T3 CO i— 

CO 4-> f- 

•m c <a 



"O CO 

1> r— ^ 

.o -^- o 

E « <a 

O V. i— 
UHCD 



C 

o 

> 

0) 

>>Q 

CO 

f— sz 

*. +■> 
o s- 

3 O 

CO z 



CO 

i. 
e 

CO 



CO 

3E 



CO 

X 
CO 



CO 

s- 

2 



OO 






CO 
«/> 

s- 
<a 
o 
o 



CO 






CO 



CO 



c 



4-> 



3 
U 



<T3 
U 



CO 

3 



O 
C 

CO 

V. 
CO 

a 

CO 



T3 
C 

ia 

CO 
Ol 

i— </> 

4-> ^r 
o I— 
t— -K 



223 



Glazed Redware, but were differentiated on the basis of decorative technique. 
Trailed Redware had a white slip applied to the surface in lines. Funnelled 
Redware had reddish brown and white slips applied simultaneously in wavy 
parallel lines. Both of these last two types were then covered with a clear 
glaze, were thick, and had forms restricted to large bowls or large round bot- 
tomed plates often with a "pie crust" rim. On Funnelled Redware the white 
slip tended to pop off, leaving the reddish brown lines in place. This was 
probably due to the better adhesion and expansion properties of the reddish 
brown slip. 

Slipwares have been isolated from redwares on the basis of a buff paste, 
forms, and surface treatment. It was also found that once the slipwares were 
separated from the redwares and divided into types, these types seriate in 
time, which suggested their potential usefulness as relative dating tools. 
Slipwares were divided into two groups, those with a clear glaze and those 
with a yellow tinted glaze. The clear glazed group was further divided into 
three types, Plain Clear Glazed, Combed Clear Glazed, and Trailed Clear 
Glazed. In addition to having a buff paste and clear glaze, these types also 
shared similar bowl and cup forms. The major difference between them was 
surface treatment. Plain Clear Glazed included all those sherds which could 
not be definitely placed into either Combed or Trailed. Combed had a reddish 
brown slip covering the vessel in areas where decoration was desired. The 
entire exterior (and occasionally interior) surface was covered with a white 
slip and then scratched or combed to expose the reddish brown slip. Trailed 
had the white slip applied first and the reddish brown slip applied on top of 
it in lines and/or dots. These two methods of decoration were never mixed on 
the same vessel and were usually readily identifiable. 

The second or tinted glaze slipware group was divided into four types; these 
were Plain Tinted Glaze, Combed Tinted Glaze, Trailed Tinted Glaze, and Black 
and Trailed Tinted Glaze. Plain Tinted Glaze Slipware had a yellow cast to 
an otherwise clear, probably lead, glaze. All of the undecorated sherds 
which could not be put into the Combed or Trailed types were included in this 
type. Combed and Trailed Tinted Glaze Slipware were manufactured and decor- 
ated in the same manner as the corresponding clear glazed types. The only 
discernible difference was the yellow tinted glaze. Black and Trailed Tinted 
Glaze also had the yellowish glaze, but was otherwise unlike any of the other 
types in surface finish. First, a black slip (actually a very dark reddish 
brown slip) was applied to the buff paste, then a white slip was trailed in 
narrow (+ 3mm wide) lines or dots on the dark surface. The whole was then 
covered with a tinted glaze. 

Figure 72 shows a seriation of the Clear and Tinted Glaze types using the 
relative frequencies of the types at four loci within the sites studied. 
Only the sherds from block excavation at Structures 76A and 76B were used 
from Site 38BK76. The percentages from 38BK75 represent material from the 
excavated block, and those from Site 38BK245 represent all material collected 
from the stripped area at that site. The chart clearly separates the clear 
glazed material (on the left) from the yellow tinted material (on the right) 
and correctly aligns the sites 1n time according to independent dating cri- 
teria, including MCDs and historical research. There was, of course, some 
overlap of the types as would be expected of sites which overlap in time. 
Two types, Plain Clear Glazed and Black and Trailed Tinted Glaze, did not 



224 









ssb 




M O 

0D<m 



25S 

<2a 



Q.O 




O 





O 






O 






C JZ 






■<- 






•<- s_ 






4-> S_ 






T3 3 






•.- <J 






S_ 






OJ "O 


00 




00 C 


c 


CM 


f0 





r- 


CD 


•r— 




s- c 


■t-J 


UJ 


fo <a 


13 


ac 


3 -c 


4-J 


ZD 


Q. cn 


C 


c; 


•<- 3 


<T3 


►— 


1 — fO 


1— 


Ll_ 


00 >■ 


a. 



o" 
o 



55 

o 



o- 1 



CD 
00 



OD 
(0 

OD 
00 



< 

<x> 

CD 
00 



IT) 
<t 

CM 

CD 

00 
fO 



225 



overlap and appeared to be the clearest time markers. The occurrence of only 
two combed Clear Glazed sherds at 38BK245 and no such sherds at Structure 76A 
at 38BK76 indicated that this type may also be a good indicator of later 
occupations. 

The earlier occurrence of the yellow tinted glaze and the later clear glaze 
may show an improvement in glaze types used by one or more potteries which is 
distinguishable archaeologically. The independent relative dating of the 
seriation supports this hypothesis. 

Unglazed coarse earthenware of local manufacture has usually been defined as 
prehistoric on primarily prehistoric sites and Colono-Indian on historic 
sites (Fairbanks 1962, South 1974:181, Noel Hume 1962, Ferguson 1977). As 
Ferguson and others have pointed out, there is reason to believe that not all 
Colono-Indian is, indeed, Indian. If it were possible to differentiate be- 
tween Indian- and slave-made pottery at Yaughan and Curriboo plantations, and 
if this distinction could be extended to other sites, new avenues would be 
opened for exploring the economics and social organization of slavery in the 
American South. Once a differentiation could be made between Indian- and 
slave-made ceramics, new questions could be asked of the archaeological rec- 
ord. The presence or absence of slave or Indian ceramics could be examined 
not only in the older eastern seaboard areas, but also in other areas of the 
South where contacts with African slaves and native Indian groups were main- 
tained. Some questions that might be explored are: 

Why are there few examples of Indian- or slave-made ceramics after the 1830s? 
Why are there few examples of slave-made ceramics in the southeastern Pied- 
mont and Sea Island areas of Georgia? What do the relative frequencies of 
Indian- or slave-made ceramics in relation to each other and to nonlocal 
ceramics tell us about the economic and social conditions of the makers and 
users of the ceramics? What differences are there between slave-made cer- 
amics in Virginia as opposed to South Carolina in attributes such as form, 
decoration, method of manufacture, and function? And finally, where do the 
attribute modes found in slave-made ceramics come from: Africa? The West 
Indies? Only a few select areas? Are they a mixture of many different 
ceramic traditions, including native American? 

The literature and, in a few cases, visual inspection of other collections 
indicate that probably both free Indians and slaves made ceramics for their 
own use and for trade and sale. For a more complete discussion of this 
subject, Ferguson's (1977) paper on the "'Afro' in Colono-Indian Pottery" 
should be consulted. Other researchers not referenced by Ferguson indicate 
the presence of unglazed ceramic forms similar to our material from slave 
sites on Barbados and Jamaica (Handler and Lange 1978, Ebanks 1974, and 
Matthewson 1973); St. Kitts, St. Thomas, St. John, St. Vincent, St. Martin, 
and St. Croix (Gartley 1979:47-61); Antigua (Handler 1964); in Virginia 
(Henry 1980); and Berkeley County, South Carolina (Anthony 1979). 



226 



Lei and Ferguson (personal communication 1980-81) is studying the regional 
variability of Indian- and slave-made ceramic attributes from Virginia to 
Georgia. More work is needed and will be done in the future. We feel that 
our contribution to the field can be a descriptive analysis of Indian- and 
slave-made ceramics and how these have shed light on our study of slavery at 
two plantations in South Carolina. With a collection of over 21,000 coarse 
unglazed sherds from both plantations, we feel that the contribution can be 
significant. 

Even as work progressed during fieldwork, certain qualities of the unglazed 

earthenwares became obvious; these were their vast numbers, the lack of any 

decoration corresponding to the Woodland ceramics at 38BK76, the presence of 
red painted lines on the finer buff colored pieces, and the incidence of 

small jars or pots, especially evident at 38BK76. Perhaps as a result of 

these predispositions from the field the most interesting results of the 
analysis revolved around these same themes. 

Since previous descriptions of Colonoware ceramics in the area (Drucker and 
Anthony 1979, and Anthony 1979, among others) were based upon relatively 
small samples and many were in the process of being written while our anal- 
ysis was being conducted, we were unable to rely on published descriptions to 
any great extent. Our objective was to develop a typology based on the 
type-variety system (Matheny 1970; Smith, Willey, and Gifford 1960; Sabloff 
and Smith 1969; Haberland 1963; Wheaton 1976; and Ball 1973). The time and 
budget allowed for such a study was limited, however, and the results are not 
as detailed as we would have liked. 

The analysis began by inspecting every sherd and grouping a large sample of 
similar sherds into ceramic units on the basis of paste color and texture, 
nonplastics, interior and exterior finish, surface color, and form. It be- 
came readily apparent that the variation in firing control and clay sources 
of the pottery was such that only two basic types could be consistently de- 
tected; these were a thicker, poorly fired, poorly manufactured type and a 
thinner, better fired and manufactured type. Internally, the thinner was 
more consistent in clay and nonplastics, as well as in color, than the thick- 
er type. Further examination revealed that there were several mutually ex- 
clusive sets of attributes separating the two types, and that the thicker 
type could be further broken down into two varieties, tooled and smoothed. 
The attributes were thickness, form, surface finish and color, decoration, 
and to a lesser extent, method of manufacture. The salient sorting criteria 
are given below and vessel forms are presented in Figures 73 and 74. We have 
chosen, for reasons discussed below, to call the thicker type at Yaughan and 
Curriboo, Colono, and the thinner type, Catawba. 

The Colono and Catawba rims were inspected to determine whether they were 
from one of the following rim/vessel types (Shepard 1965): open-incurving, 
closed-incurving, outsl oping, outcurving, and unidentifiable. The lips were 
also inspected for a variety of attributes including rounded, flattened, 
folded, tapered, etc. Although the lip attributes were used in the ceramic 
typology, they did not provide useful information for form studies. 



227 




TOOLED JAR 



SMOOTHED JAR, RIM INCOMPLETE 









^Hj 


Ky«v^ >f / 'JB^155JctI«m 






vS**"'**?** *^fB3f- \ 












iSck.^ J 






nnjL 












Sgfc.''* ' v&/ 






w^^ " *~ jHF 




^! 









TOOLED PEDESTAL JAR 



TOOLED OPEN INCURVING 80WL 



FIGURE 73 
Colono Pots 



228 



<r 

= CE 

«C < 



— 03 



LU 

<r 

< 

< 

CD 

< 

< 




0> 
CM 



0. -I 

o* 

JO 
CO o 



k 



r 



JO 

3 
re 
+J 

ns 

C_) 

-a 
«r c CI 

N H! S 

i. 

LU O O 

=> o 

CD i— = 
■— O •— 
u. <_> e; 



JE 






JQC 



tr 
< 




co 

o 

a. 

<r 

LU 

CD 

2 
< 
X 

u 



cv. <*• 



r 






CE 



<o 

-j 

O 
CD 

O 

z 

> 
tr 

15 
U 

z 
z 

LU 

a. 
o 



vlv 



v^ 














229 



COLONO 



CATAWBA 



Thickness Average .725 cm thick up to 
very uneven on indi- 
vidual vessels and even sin- 
gle sherds. 

Form Generally open incurving bowls 
and small flared mouth jars, 
lips were crudely rounded, or 
flattened with a finger or 
stick. 



Average +.5 cm thick; 1.1 cm, 
regular and even. 



Generally straight sided, open, 
outflaring bowls, and small well 
made jars, lips were tapered and 
well finished. 



Body 



Surface 



Wide variation in size, 
amount and type of non- 
plastics, generally various 
water-washed sands, oxida- 
tion was usually not complete, 
leaving a dark core. 

Ranged from crudely smoothed 
to polished with obvious evi- 
dence of the polishing tool, 
generally interiors of bowls 
and exteriors of jars were 
polished, color ranged from 
black to dark brown to red- 
dish orange, great variation 
on individual vessels and 
sherds. 



Decoration .3% had decoration on inte- 
rior of bowls including pre- 
firing notched rims, reed 
punctate, thimble impressed, 
incised lines; post firing 
incision in the form of a 
cross in a square and a cir- 
cle occured on the interior 
bottoms of a few bowls (Figure 
75). 

Method of Bases occasionally coil made 
Manufac- and body was hand modelled, 
ture poor control over firing 

temperature and firing time, 
handles appeared to be at- 
tached to the surface of the 
vessel (Figure 76). 



Limited variety of nonplastics, 
generally fine particle size 
and completely oxidized or com- 
pletely reduced. 



Usually highly polished on in- 
terior and exterior of bowls 
and wide mouthed jars, polish 
marks were often evident, color 
rangeds from black to gray to 
buff, little variation on in- 
dividual sherds, some vessels 
were intentionally reduced. 



3.5% of Catawba had undulating 
"day-glo" red painted lines on 
the exterior of jars and the 
interior of bowls applied after 
preliminary or final firing of 
the vessel; occasionally red dots 
were placed around the undulating 
line, or around small regular 
facets taken out of the interior 
lip; or both. 

Evidence supports hand modelling 
but sample is too small for de- 
finite conclusions, firing temp- 
erature and time were well con- 
trolled, reduction when it occurs 
was intentional, handles had plugs 
on the end which were inserted in 
the wall and smoothed from the 
inside. 




230 
INCISED 

TOP- 3 INCISED 
BOTTOM -3 INCISED 
RIGHT -"PEELED" SURFACE ON 
PUNCTATE SHERD 




PUNCTATE 
LEFT-PIPESTEM PUNCTATE, 

SMALL REED PUNCTATE 
CENTER -LARGE REED PUNCTATE 
RIGHT TOP-PIPESTEM PUNCTATE, 

"HORSESHOE" PUNCTATE 
RIGHT BOTTOM- SMALL REEO 
PUNCTATE 




SMOOTHED 

TOP- INCURVING RIM, 

OUTCURVING JAR RIM 
BOTTOM - 2 OUTCURVING JAR RIMS 



FIGURE 75 

Decorated, Tooled, 

and Smoothed Colono Sherds 




231 
BASES 

TOP- RING BASES 
BOTTOM - PEDESTAL BASES 




APPENOAGES 
TOP - HANOLES 
BOTTOM -LIO HANOLE, 

SUPPORT OR LUG HANOLI, 

HANOLE 




CENTIMETERS 



OBJECTS 

TOP -FIGURE BASE (?), 

PERFORATED STRAINER , 
FINGER IMPRESSED CLAY LUMP, 
PUNCTATED CLAY LUMP 
BOTTOM -3 MAR8LES, 

PIPE BOWL BASE, 
2 PIPE STEMS 



FIGURE 76 
Colono Bases 
and Objects 



Appendages , 



232 



Colono had two varieties, Smoothed and Tooled. As noted above, the surface 
finish on Colono varied widely. Smoothed Colono included those sherds with a 
poorly finished or wiped surface still showing evidence of finger impressions 
and a very uneven wall thickness on single vessels and sherds (Figure 75). 
The one reconstructed vessel of this variety was broken into much smaller 
fragments than any of the reconstructed vessels of the Tooled variety. This 
was probably because Smoothed sherds generally had more nonplastics than 
Tooled and were more friable. Possibly four of the Smoothed vessels at 
38BK76 (minimum vessel counts were impossible within the project schedule) 
were very small jars, probably no more than a few inches high and half as 
wide (Figure 75). These appeared to have been made by or for children. 
Tooled Colono ranged from a poorly polished to fairly well polished surface, 
although the polishing marks usually did not overlap. On a few distinctive 
sherds the exterior surface appeared to have been cut or shaved, leaving 
large irregular facets. Although the range of nonplastics and thickness 
covered those of Smoothed Colono, the general trend was for Tooled to have 
fewer nonplastics and a more regular wall thickness on single vessels and 
sherds. Once these type descriptions were developed, all Colono and Catawba 
sherds were laid out together and classified at the same time. 

Spalling on pots was also noted. It seemed reasonable that such pots would 
have been discarded near where they were made if they showed no evidence of 
use. Unfortunately, some of the spall ed pots did show evidence of having 
been cooked in, leaving a charred residue. Some sherds exhibited peeling on 
the exterior surface. This peeling of the top millimeter or two is not to be 
confused with spalling. The line between peeled and polished surfaces is 
straight, unlike the edges of a spall scar. The line does not follow the 
natural curve of the vessel as would be expected if the vessel were accident- 
ally hit by a hard object. The overall shallowness of the peeled area and 
the straight line between peeled and polished surfaces suggests that a corro- 
sive liquid was allowed to stand in the pot on a regular basis. This liquid 
seeped through the pot and eventually caused the exterior surface to flake 
off where it had alternately been wet and dry. Similar flaking is seen on 
sherds and vessels found in heavily saline soils in other areas. The unknown 
liquid may have been salty or high in other minerals. These sherds may rep- 
resent chamber pots, and one rim sherd does have the appearance of a chamber 
pot rim (Figure 74). 

Perhaps the most important find associated with the local manufacture of cer- 
amics were two incompletely fired Colono sherds found at Site 38BK245. One 
of these could be termed unfired since it was so soft that a damp brush would 
have destroyed it during cleaning. Clay objects made of Colono clays were 
also found at the sites. These are listed in Table 32 and illustrated in 
Figure 76. 



233 



OBJECT 



Table 32. Colono Objects 
38BK75 38BK76 



38BK245 TOTAL 



Pipe Parts 

Handle Parts 

Marbles 

Strainer Parts 

Handle or Support Parts 

Lid Knob 

*Misc. Objects 

Unfired Colono Sherds 



8 


4 


12 


15 


3 


18 


4 




5 


3 




3 


1 




1 


1 




1 


4 




4 




2 


2 



Total 



36 



46 



*Misc. Objects - finger marked lump of fired clay, lump of fired clay with 
many holes, small thin striated object, and flat, flaring object worn on 
edges. 



It should be noted that Colono pipe parts included short stems and bowls, and 
that these were fairly well made with a fine paste. Handles were crudely 
made and were oval to slightly flattened in cross section (Figure 76). These 
were assumed to be for jars or pots. None showed evidence of being inserted 
into the vessel wall like those of Catawba (Figure 77). Some may have been 
lug handles, but this was undeterminable in our sample. The objects identi- 
fied as strainers (Figure 76) were flat thin pieces of fired clay, always 
with broken edges and holes perforated through them. These were crudely 
smoothed in two cases and well smoothed, but not polished, in one case. They 
may have represented strainers in teapots (Lewis and Haskell 1980:104) or 
simply strainers in the bottoms of bowls, as no clear indication of Colono 
teapots were found. One straight handle or possibly a support was found 
which had been faceted much like some pieces of Tooled Colono.* Four miscel- 
laneous objects which have unknown functions were also found (Figure 76). A 
lump of fired clay which had been squeezed in a hand leaving finger marks may 
have represented a sample piece used to test the clay for firing properties. 
Another was a lump of clay with many deep holes gouged in it which morphologi- 
cally resembles a child's attempt at making a pencil holder. A flat piece 
with ground edges resembled the base of a skirted figurine. The final object 
was very small and had incised or impressed parallel lines on it which may 
simply be from being pressed against a reed or piece of grass. These last 
four objects were probably not the sort of thing that would be marketable, 
which would tend to indicate that they were made on the site rather than 
being traded in from outside. The generally inconsistent quality of Colono 
would tend to make it less saleable than Catawba. 



234 



Our Catawba closely resembles the plain modern Catawba ceramics on display at 
the Charleston Museum in color and surface treatment. Colono and Catawba cer- 
amics were discussed with Elaine Herold and Alan Lise during a visit to the 
museum in the fall of 1980. In the course of the discussion, Mr. Lise pro- 
duced from storage a small, handled Catawba pitcher with "day-glo" red paint- 
ed, wavy line and dot decoration (Figure 78). Although it was completely 
reduced, unlike the majority of our decorated Catawba material, the highly 
polished surface, thin walls, and decoration were unmistakable. This pitcher 
(Charleston Museum accession #ETN124) had been donated by David Doar, the 
great grandson of Dr. Samuel Cordes who, according to tradition, had bought 
it from a Catawba woman in St. Stephens in 1805. To our knowledge, this is 
the earliest attributable piece of Catawba on record and was bought by a man 
who at one time owned an interest in and occupied Yaughan plantation. 

At all three sites and especially at the excavated portions of Site 38BK76, 
large deep features filled with alluviated soil and refuse had been excavated 
by the inhabitants. These features represented either intentional trash pits 
or clay extraction pits for construction or ceramic clay. That the slaves 
would intentionally dig trash pits while the more typical English pattern of 
refuse disposal seems to have been to leave trash on the surface (South 
1977a: 47) seems contradictory. The apparent squalor in which the slaves at 
Yaughan and Curriboo were forced to live, as illustrated by other data, in- 
cluding their artifact patterns, housing, and subsistence, also seems to 
contradict intentional trash pits in most cases. 



*In France, the Huguenots are known for, among other things, a marmite 
huguenote . According to the Nouveau Larousse II lustre (n.d.) Volume 5, the 
marmite "Ruguenote is a "marmite de terre sans pied ou avec des pieds, tres 
bas. [Les huguenots, dit-on, se faissaient apporter leurs repas dans ces 
recipients, le jours d'assemblee et de preche). Petit fourneau avec la 
marmite qu'il recoit". Loosely translated, the text reads, "earthenware pot 
with or without very short supports (feet). (The Huguenots, traditionally, 
carried their meals 1n these recipients on assembly and worship days). Small 
brazier with the pot it holds." If the support mentioned here is from a 
marmite huguenote or a copy of one, then there is a suggestion that some 
Huguenot culture may have been retained after 1740 and may be reflected in 
slave culture. 



235 




en 

Z 
UJ 

o 

< 
cr 
u. 

UJ 



o 

<r 




1 


1 


■i ui 


t- 


UI 


II 


UI 


o 


1 


? 


1 



«*- 

s- .*-> 
OJ s_ 

j= < 

O Q. 

"O o 
c - 

UJ XI OJ 

3 «: u 
UI yj +j 4_j 

-J _ rg m- 

I— L.U ^ 

o 

CD 

"ui 

b. 

tr 

UI 
_0L 

en 
en 

< 

o 

DC 

< 
UI 




en 

o 

< 
<r 
ui 
o 

< 

< 
< 




cn»- 
UJ.Z 

t°- 

Oq 
032 
UJ << 

2<n 

UJq: 

a:_ 

-IUJ 

O-J 



Z 

o 
a. 



< 
en 



z z 

o o 
a. a. 

en <n 
en en 
< < 

o o 



236 




Photograph taken by 
Charleston Museum 

Approximately 3/4 scale 



FIGURE 78 
Catawba Pitcher 



237 



The following experiment was conducted in order to determine whether the 
clays at the sites could have been used for manufacturing Colono pottery. 
Nineteen soil samples were used from Sites 38BK76 and 38BK245. The samples 
from Site 38BK75 and naturally occurring soils at the other sites were fired 
in a kiln. All the samples, whether they were originally red or gray, became 
redder after being fired. The samples were fired for over eight hours at 
cone 06, which is undoubtedly a higher temperature and longer firing time 
than that undergone by colonoware. The oxidizing atmosphere assured complete 
oxidation of the samples which could be compared to oxidized pieces of 
Colono. Incomplete oxidation cannot allow objective comparisons of color for 
various reasons (Shepard 1964). Unfired clays changed from 10YR or 2.5Y to 
5YR or even 2.5YR when fired; many also became only slightly lighter, but 
with more chroma, e.g. 7/2, 6/2, or 5/2 to 6/8 or 7/6. None of the fired 
samples were gray, although several of the unfired samples, especially from 
Structure 245B, were gray before firing. 

Potting capability is a subjective term to denote those fired clays which 
appeared as hard or harder than Colono, did not crumble when broken, had the 
same or a finer texture than Colono in cross-section and did not crack in 
firing. Table 33 gives data on the 19 samples. 







TABLE 33. 


Potential Potti 


ing Clays 




Feature 


Soil 


Original 


Fired 


Context 


Potting 




Texture 


Color 


Color 




Capability 


76I-F24 


C 


10YR 5/3 


2.5YR 


6/8 


trench 


poor 


245K-F1-2 


C 


7.5YR 6/8 


2.5YR 


5/8 


kiln 


excellent 


245C-F4-1 


SC 


10YR 7/6 


2.5YR 


5-6/8 


hearthpad 


poor 


245B-F11-7 


c 


2.5YR 6/4 


5YR 


7-6/6 


trench 


good 


245B-F11-8 


c 


2.5YR 6/2 


5YR 


7/6 


trench 


good 


245B-F11-11 


c 


10YR 5/2 


5YR 


7-6/6 


trench 


excellent 


245B-F17 


c 


10YR 5/2 


5YR 


7/6 


trench 


excellent . 


245B-F11-18 


c 


2.5YR 7/2 


5YR 


7/6 


trench 


good 


245B-F11-23 


sc 


10YR 7-6/3 


5YR 


7/6 


trench 


good 


245B-F11-20 


c 


10YR 6/4 


5YR 


7/6 


trench 


good 


245B-F11-26 


c 


10YR 6/3 


5YR 


7/6 


trench 


poor 


245B-F11-29 


c 


10YR 6/2 


5YR 


7/6 


trench 


good 


245-F18-2 


c 


10YR 5/3 


5YR 


7/6 


posthole 


good 


245-F31-2 


c 


10YR 6/4 


5YR 


7/6 


posthole 


good 


245-F38 


c 


2.5Y 5/2 


5YR 


7-6/6 


posthole 


good 


245-F41 


c 


10YR 7/3 


5YR 


7/6 


posthole 


good 


245-F57-1 


c 


2.5Y 6/3 


5YR 


7/6 


posthole 


good 


245-F58-1 


c 


10YR 6/4 


5YR 


7/6 


posthole 


good 


245-F61 


c 


10YR 6/4 


5Yr 


7/6 


trash pit 


good 





Despite the limitations imposed on the selection of samples and their rela- 
tively small number, it can be concluded that nearby, naturally occurring 
clays, which are virtually identical to Colonoware in texture and oxidized 
color, were available to the plantation inhabitants. 



238 



The lack of kilns, either of the subterranean variety or above ground type, 
and a similar lack of heavy charcoal concentrations at either plantation 
seems to argue against on-site manufacture of Colono. This may be a result 
of western preconceptions concerning pottery making. In the Southwest and in 
Mexico, among other places, ceramics are often fired above ground when only 
small quantities are required. This was also true of eighteenth century 
Barbados (Handler and Lange 1978:139-144). Charles Counts (personal commun- 
ication 1981), a specialist in modern Yoruba pottery, noted similarities 
between our material and African forms, as did Henry Drewel (personal com- 
munication 1980). Counts further noted that a typical firing technique of 
local potters in Nigeria (of whom he has met well over 100) is to simply pile 
the greenware with the fuel on the surface of the ground. This fuel is often 
grass. Fagg and Picton (1978:11) also illustrate this firing technique in 
Nigeria. 

The poorly fired pots of Colono usually had firing clouds and dark cores: 
these attributes, combined with the underfired or unfired sherds from Site 
38BK245, would imply a short, poorly controlled firing as suggested by Counts 
and Fagg and Picton. Similar firing clouds and cores have also been produced 
by one of the authors using an open air, above ground firing technique. The 
absence of kilns cannot only be explained from ceramic evidence, but, consi- 
dering the background of the potters and their economic and social condition, 
one should not expect to find formal kilns. A similar lack of heavy charcoal 
concentrations, indicating surface firing, should also be expected since re- 
latively few vessels were probably made over the lifetime of the sites. As- 
suming an average total of 40 slaves at the peak of Colono production between 
1740 and 1780 with a requirement of one bowl per person and three cooking 
pots for every five persons, and allowing for an average six month lifetime 
for the pots, approximately 128 pots would be needed every year. This could 
have been covered easily by six relatively small firings of 21 pots. The 
total result would be 240 firings of 21 pots each, over a 40 year period. 
Site 38BK76 covered approximately three acres, not including peripheral 
areas. Two hundred and forty firings placed in various parts of three or 
more acres over a period of 40 years could not be expected to leave heavy 
charcoal concentrations, especially if the fuel used was grass, the ashes of 
which would be blown around by winds and washed away in storms. In all stra- 
tigraphies, there were quantities of charcoal flecks fairly evenly mixed 
throughout the soils. This was at first thought to represent evidence of 
burning off of fields, but may have also represented wind blown deposits of 
ash from Colono firings and cooking fires. 

The rounded bottoms of both the Colono pots and the bowls with practically no 
handles and blackened bottoms suggests that the cooking vessels were placed 
directly on the fire. Handler and Lange (1978:54) point out that slaves on 
Barbados preferred cooking in family groups and did so without permanent 
hearths or fireplaces. This may also be a common practice in Africa (Fagg 
and Picton 1978:17). In Mesoamerica and Barbados (Handler and Lange 
1978:54), several, usually three, stones were placed on the ground upon which 
the cooking pot was set. If this was the case at Curriboo and Yaughan, then 
this may explain the absence of fireplaces and the presence of only one 
hearth (plus two possible hearths) at the plantations. 



239 



In Colonel Thomas Cordes' (16977-1748/9) inventory at his death, his most 
expensive slave, at 400 L, was "1 Negro Man Named Potter" (or Porter) 
(Inventory of Col. Thomas Cordes 1748-1751). Since slaves with crafts were 
more expensive and slaves were occasionally named for their skill, there is a 
possibility that this slave represented a black potter or porter at either 
Curriboo or one of Thomas Cordes' holdings in St. John's Parish, Berkeley. 
That slaves traded other objects which they made, such as boats, chairs, 
baskets, and food, etc. to owners on their own and other plantations is 
established by historical research. There are no records of slaves trading 
Colono to owners, although there are records of owners buying Catawba cer- 
amics. The similarities in the Colono at Yaughan and Curriboo, making them 
virtually indistinguishable, implies contact or even trade between the plan- 
tations. Although there was no conclusive proof of this (e.g. identical 
decoration at the two sites, signed pieces, trace mineral analysis, etc.) the 
opportunity was present and trade in other items was conducted. 

Comparisons of the Curriboo and Yaughan material with that of other areas 
produced strong similarities with that illustrated by Drucker and Anthony 
(1979) and Anthony (1979) at Spiers Landing. Unfortunately, inspection of 
their collection was not possible. The Colono material in and near 
Charleston was found to be similar, but it tended to be more consistent in 
manufacture, better finished, and clearly had several European forms, e.g., 
teapots and plates (Charleston Museum collections; Lewis and Haskell 1980). 
From this a tentative hypothesis can be made that Colono-African pottery 
along the entire eastern seaboard will fall into one of two basic form 
groups: one on inland and more or less isolated plantations and a second in 
or near cities (i.e., Rural and Urban Colono). The first will have gener- 
alized and fewer forms for local use, and the second will have more forms 
with copies of European forms and will have been used for sale, perhaps even 
to whites. 

To summarize this description of the unglazed earthenwares, several facts 
were established about such pottery from the literature cited above: 

1. Forms of Colono in the West Indies were similar or identical to those at 
Yaughan and Curriboo plantations and elsewhere. 

2. There were sites with Colono, notably in the West Indies, where there was 
no native ceramic tradition before the introduction of African slaves. 

3. The European cultures at these same sites ranged from English to Dutch to 
French and other cultural backgrounds. 

4. Native American Indians were making pottery for trade or sale to slaves 
and slave owners alike in Coastal South Carolina and elsewhere. 

5. The Native American pottery varied from region to region and may have 
been attributable to known tribes. 

6. Ceramics similar to Colono appeared late in the seventeenth century and 
died out 1n the first half of the nineteenth century. 

7. Ceramics similar to Catawba appeared at least by the end of the eight- 
eenth century in South Carolina and are still being made today. 



240 



From these observations, several conclusions can be drawn. The makers of 
Colono were either from the same or a similar cultural background or were in 
fairly constant contact with each other during the Colonial Period. The 
possible candidates for potters included Native Americans, one of several 
European groups, or African slaves. 

In order to explain these facts, one of several conditions must obtain. 

1. Native Americans with a common ceramic tradition shifted to making Colono 
in the late seventeenth century and proceeded to sell or trade their wares, 
produce them, or instruct others in their manufacture over an area including 
South Carolina and the West Indies, and perhaps farther. This is unlikely 
for various reasons. Regular trade or travel on so vast a scale across wide 
expanses of water by a group of Native Americans (outside of Mesoamerica) 
with a common ceramic tradition, to say nothing of interference by European 
interests in such movement, were highly improbable coincidences. 

2. The second possible condition would involve a single European group or 
groups with a common cottage pottery industry, and with access to other 
European markets, making and trading Colono. The restrictions on trade 
between colonies of different European powers, however, would indicate that 
if such trade went on the risks to circumvent trade restrictions would be too 
great to justify trade in Colono when European ceramics were more or less 
readily available. 

3. The third possible condition involves African slaves. In this case, 
Africans (including African potters) with similar ceramic traditions settled 
in the West Indies and on the mainland of North America. Using the resources 
at hand and being in poor economic straits, they produced colonoware. This 
would explain the similarities within colonoware ceramics, as well as the 
minor differences, the date of its appearance, the fact that it appears on 
sites with no prior native ceramic tradition, and that it occurs in areas 
controlled by various European powers. African slaves are the only common 
link tying together all of the known facts from the literature concerning 
this pottery at the present time. 

Included within this last hypothesis is the possibility that native American 
or Indian slaves produced or instigated the production of slave-made pottery 
rather than African slaves. We feel this is highly unlikely for the fol- 
lowing reasons, summarized from the historic research and archaeological 
evidence: 

1. According to current historical investigations on the subject, Indian 
slaves were ne^er more than 25-30 percent of the slave population of South 
Carolina. 

2. Indian slavery peaked by 1710 or so and was in rapid decline towards in- 
significance by 1730. 

3. The impact of Indian slaves on other aspects of plantation culture has 
been shown to have been insignificant. 



241 



4. The slave population began to show significant growth after 1711 with the 
introduction of large numbers of imported African slaves. 

5. Yaughan and Curriboo were settled after 1740 and have no definite Indian 
slaves appearing in the records. 

6. The Santee River area was virtually depopulated of free Indians before 
European settlement of the region. 

7. The Colono at Yaughan and Curriboo appears to have much stronger ties to 
with strictly slave sites than with known Colono-Indian ceramics. 

It has been concluded on a general level that some slaves made and used their 
own ceramics and that Indian slaves probably played little or no role in 
Colono manufacture. On specific regional levels, it is also apparent that 
different free Indian groups made ceramics for their own use and for trade or 
sale. On a local level, two types of ceramics, Colono and Catawba, occurred 
on the plantations. Based on evidence presented in the type description and 
the general conclusions reached from the ceramic and historical literature, 
It can be concluded that Colono was made by and for slaves and Catawba was 
made by Indians for sale or trade. 

In summary, the unglazed earthenwares at Yaughan and Curriboo were divided 
into two types, Colono and Catawba. Colono was divided into two varieties, 
Tooled and Smoothed. It consisted mainly of bowls and pots, but was also 
found in the form of pipes, marbles, and other objects. Catawba showed af- 
finities to attributable Catawba Indian ceramics, while the evidence suggests 
that Colono was made on the plantations. 

With this description, the unglazed earthenware ceramics can be analyzed in 
comparison to other artifact categories which may shed light on the character 
of the occupations at Curriboo and Yaughan. Slave-made Colono was the major- 
ity type at both plantations and all three sites. Not only did it make up 
over 85 percent of the unglazed and glazed ceramics (Table 34), but it made 
up over 66 percent of all artifacts at 38BK76 and 56 percent of all artifacts 
at 38BK245. At Site 388K75, Colono made up 38 percent of the total artifact 
assemblage. Not only was Colono the majority ceramic type, it was the 
majority artifact type at the earlier sites and made up a substantial 
percentage of the artifacts at a later site. 

Table 34 
Colono/Catawba and Nonlocal Ceramics 



Colono 



Catawba 



Nonlocal 



Total 



38BK75 


2545 


65.90 


295 


7.64 


1022 


26.46 


3862 


38BK76 


15043 


89.48 


141 


.84 


1627 


9.68 


16811 


38BK245 


3316 


87.77 


17 


.45 


445 


11.78 


3778 


TOTAL 


20904 


85.49 


453 


1.85 


3095 


12.65 


24451 



242 



Table 35 presents the quantities and relative frequencies of Colono, Catawba, 
and nonlocal ceramics of 38BK75, 38BK245, and 38BK76, Structures 76A and 76B. 
The structures at 388K76 were used as separate sites or units since both had 
been hand excavated, were fairly far apart (which lessened the chance of 
contamination), had sizeable quantities of artifacts, and provided more sites 
or units for seriation. Figure 79 presents in graphic form the relative 
frequencies presented in Table 35. It can be readily noted that Colono 
decreased from 38BK76A to 38BK75, and that there were more dramatic increases 
in Catawba and nonlocal ceramics. Generally, these trends follow time from 
early to late. That Site 38BK245 1s out of sequence with respect to time is 
considered in some detail in Chapter XII. The main concern here is with the 
general trend of decreasing Colono and increasing amounts of Catawba and 
especially nonlocal ceramics over time. 

The rim/vessel forms used in the following comparisons corresponded to the 
restorable vessels and to forms illustrated in the literature. The open 
incurving form equates with rounded bowls, closed incurving with deep rounded 
bowls; outcurving with cooking pots and jars, and outsloping with straight- 
sided flat bottomed bowls as found in Catawba and very infrequently in 
Colono. For the purposes of comparison with nonlocal ceramics, these form 
types were collapsed into bowls and cooking/ storage pots. Nonlocal ceramic 
rims were divided into flatware, bowls/cups (to include clearly defined bowls 
and cups as well as vessels which could be either), and other forms such as 
storage or cooking vessels. Nonlocal rims represented 13 percent of the 
total nonlocal ceramics on an average. Colono rims represented slightly less 
of the total Colono sherds (10 percent), but these percentages were close 
enough so that comparisons of rims could be assumed to represent the two 
populations of sherds. 

The relative proportions of Colono bowls and pots change with respect to each 
other and to the wider variety of forms in nonlocal ceramics, especially 
bowls and plates. Statistically, these shifts have proven significant. This 
discussion of form is based upon analysis of all Colono, Catawba, and non- 
local rim sherds. A minimum vessel count of the nonlocal ceramics was com- 
pleted but could not be compared to Colono since a minimum vessel count of 
Colono was not conducted. This was because of the irregular nature of Colono 
which made clear identification of minimum vessels difficult and because of 
limited time. 

Table 36 presents the various major form categories within Colono and non- 
local ceramics. Figure 80 graphically represents the relative frequencies 
given in Table 36. The position of Site 38BK245 with respect to time is 
better illustrated here than in Figure 79, although it is very similar to 
38BK76B. Essentially, the sites are seriated from the earliest at the bottom 
to the latest at the top of the figure. 

Correlation coefficients were taken of pairs of forms to determine if the 
trends noted in Figure 80 were strong and whether or not they were sig- 
nificant. Comparing the following pairs of form categories produced the 
accompanying r values (Table 37). 



TABLE 35. Nonlocal and Local Ceramic Totals 



243 





Col 


ono 


Catawba 


Non- 


local 


Total 


38BK75 


2545 


65.90 


295 


7.64 


1022 


26.46 


3862 


38BK76B 


6762 


89.71 


85 


1.13 


691 


9.17 


7538 


38BK245 


3316 


87.77 


17 


.45 


445 


11.78 


3778 


38BK76A 


4586 


97.76 


32 


.68 


173 


3.69 


4691 



244 



< 
u 

o 



< 

IS 

! 




O 

Z 

o 

-1 
o 
o 



C r- 
O H3 
u 
"O o 
<U — I 

CQ C 
<0 

r*. o i— 

OS fl o 
3 •*- i— 

ca s- c 

i— i d) o 



O 



52 

o-| 
to 



o J 



o 

CD 
CO 



CM 

CO 
CO 



CO 



CO 

co 



CD 

CO 
IO 




245 



to 

CD 




OD 



55 

o, 
o 



$5 



O-* 



c 
o 

"O 
01 

t/1 

(T3 

CO 

O C -M 
CO o c 
■t- 3 
LU 4-> O 
OS ru C_> 

O J- £ 
1— . d) .1- 

u-uio: 



CD 
<0 

CD 
CO 

10 



C\J 
CD 



< 

s. 

CD 

CO 

lO 



246 



TABLE 36. Ceramic Rim Counts 





Flatware 


Non-1 
Bowl 


ocal 
s/Cups 


Other 


Bowl 


Col 
s/Cups 


ono 
Cook 1 


ng/Storage 


Tota" 


38BK75 


89 


17.69 


45 


8.95 


1 


.20 


339 


67.40 


29 


5.77 


503 


38BK76B 


73 


10.41 


50 


7.13 


2 


.29 


415 


59.20 


161 


22.97 


701 


38BK245 


20 


5.90 


27 


7.96 


2 


.59 


209 


61.65 


81 


23.89 


339 


38BK76A 


3 


.76 


15 


3.79 








212 


53.54 


166 


41.92 


396 



247 



TABLE 37 

Correlation Coefficients of Pairs of Forms 

(d.f. = 2) 



Nonlocal Flatware 
vs. Nonlocal Bowls/Cups 

Nonlocal Flatware 
vs. Colono Bowls/Cups 

Nonlocal Flatware 
vs. Colono Cooking/Storage 

Nonlocal Bowls/Cups 
vs. Colono Bowls/Cups 

Nonlocal Bowls/Cups 
vs. Colono Cooking/Storage 

Colono Bowls/Cups 
vs. Colono Cooking Storage 



r = +.8447 

r = +.9038 

_r = -.9674 

r = +.9515 

r_ = -.9415 

r = -9798 



£ = .155 
£ = .086 
£ = .033 
£ = .048 
£ = .059 
p =- .020 



All of the correlations were strong and three had a probability of less than 
.05. The three significant correlations were a negative correlation of non- 
local flatware and Colono cooking/storage forms, a positive correlation of 
nonlocal and Colono bowls and cups, and a negative correlation of Colono 
bowls/cups and cooking/storage forms. The general trends in time, as shown 
in Figures 79 and 80, led to the general conclusion that Colono cooking/stor- 
age forms decreased as nonlocal flatware and Colono bowls/cups increased and 
that Colono and nonlocal bowls/cups increased over time. 

With the exception of the stripped area at 38BK245, the other areas seriated 
according to their mean ceramic dates: 38BK76A dates to 1773.4, 38BK76B to 
1787.6, and 38BK75B to 1789.8. The stripped area at 38BK245 (38BK245AB) 
dated to 1760.5 which should place it as the earliest site, earlier than 
38BK76A and 38BK76B, as is shown on Figure 72. Since the mean ceramic dates 
and documentary evidence support an early to late sequence of 38BK245AB, 
38BK76A, 38BK76B to 38BK75B, and the general trend of forms follows nearly 
the same order, these trends covary, at least generally, with time. The 
anomaly of 38BK245AB may be explained as a status rather than temporal dif- 
ference. If Site 388K245 represented slaves who were better off materially 
than those at 38BK75 and 38BK76, then the process of increasing flatware and 
bowls/cups and decreasing cooking/ storage forms should have occurred earlier 
at 38BK245, resulting in a seemingly later position in a seriation. 



248 



Chemical analysis of charred remains in cooking pots was not allowed for in 
this project. Hopefully, future researchers will deal with this material. 
However, the residual remains in the pots indicate that they were used for 
cooking and that on occasion the food was allowed to burn. Such pots could 
not be used for frying or roasting and apparently they were principally used 
for boiling soups or stews. The presence of round bottomed bowls with fairly 
steep sides, while considered to be all purpose vessels, could most easily be 
used for holding individual portions of liquid or semi-liquid foods. The 
absence of knife marks (Griffiths 1978) on the interiors, while possibly the 
result of difficulties in analysis (small sherd size, erosion, determination 
of position on vessel, etc.) appeared to indicate that the bowls were not 
used for cutting surfaces. The conclusion to be reached is that the slaves 
at Site 38BK76 in the earlier period were probably cooking stews in small 
pots and eating from bowls, and that their diet did not contain large pieces 
of meat cut with a knife and fork. Such a cooking method and its implica- 
tions for diet, although perhaps monotonous, would have been relatively easy 
for slaves working in the fields all day. It had the further advantage of 
allowing adult slaves to let the food simmer and be tended by children or the 
elderly. The small cooking pots also argue against centralized cooking since 
only small quantities could be cooked at any one time. 

Because of the durability of iron kettles, their presence in the archaeologi- 
cal record is lower than it would be if they had not been so durable and 
valuable, i.e. kettles would tend not to break in the first place and those 
that did would not break into many pieces. It is also probable that useable 
kettles were carried away when the sites were abandoned. In such a case, one 
should expect evidence of fewer kettles in the archaeological record than 
were present originally. Assuming that these two factors played equal roles 
at all sites, then the relative frequencies, supported by a fairly strong r_ 
value on Table 38, strongly suggest that kettles were replacing Colono 
cooking pots by the end of the eighteenth century. 

In summary, two types of colonoware ceramics were differentiated: these were 
Colono-African (Colono) and Colono-Indian (Catawba). There is evidence that 
Colono was produced at Yaughan and Curriboo. This evidence includes a varie- 
ty of skill levels apparent in the ceramic material, local clays which com- 
pare favorably with Colono, clay extraction pits, apparently unmarketable 
Colono items, and unfired sherds. Colono occurred in generally the same 
forms and finishes, not only at Yaughan and Curriboo, but at other sites in 
the vicinity and region. Catawba was similar to pieces directly attributable 
to modem Catawba groups and to the earliest documented Catawba pot, which 
was bought by Samuel Cordes at Yaughan in 1805. It was therefore concluded 
that Colono was made and used on the plantations and that Catawba was made by 
Indians, primarily for sale or trade. Further, the similarities between our 
material, especially in form, with ceramics in Barbados and Africa led to the 
conclusion that African slaves made Colono. 



TABLE 38. Kettle Fragments and Colono 
Cooking/Storage Forms 



249 







Kettl 


les 




Pots 


38BK75 


8 




21.62 


29 


78.38 


38BK76A 


4 




2.35 


166 


97.65 


38BK76B 


3 




1.83 


161 


98.17 


38BK245 


3 




3.57 


81 


96.43 



Total 



37 
170 
164 

84 



Correlation Coefficient 
r = -.7290 
p_ = .271 
d.f. = 2 



250 



An interesting sidelight and avenue for further research, as a result of 
differentiating Catawba and Colono, involves culture change among the Catawba 
Indians. During early contact with the Anglo-American colonists, historic 
accounts speak of treating with Indians as a separate nation. Contacts were 
primarily for fur and slave trading, and among the Catawba only the males and 
tribal leaders had much contact with Anglo-American traders and political 
leaders. This changed in the early to mid-1700s so that by 
counts of contact with the Catawba were of Indian women 
Charleston trading craft goods, including ceramics (Hudson 
This is supported archaeologically at Yaughan and Curriboo by 
of Catawba ceramics at about the same time. The question of 
is therefore raised: does the appearance of Catawba indicate 
within a native American population? 



1780, most ac- 
travelling to 
1956:144-149). 
the appearance 
culture change 
culture change 



Bottle Glass 

Most of the remainder of the material which made up the Kitchen Group was 
bottle glass. Bottle glass was divided into seven categories originally; 
these were olive wine bottle, bluish olive wine bottle, other olive bottle, 
clear, green tinted, amethyst, and "other". Only one piece of amethyst glass 
(which post-dated the slave occupation) was found at Site 38BK75 and the 
"other" category had none. For the purposes at hand, all olive green glass 
was lumped together and listed by site with the other types (Table 39). 

TABLE 39. Bottle Glass Varieties 



38BK75 



38BK76 

1 



T 



38BK245 



T 



Olive Glass 


425 


76.44 


Clear Glass 


68 


12.23 


Green Tinted Glass 


63 


11.33 



1680 85.63 
164 8.36 
118 6.01 



654 


94.92 


23 


3.34 


12 


1.74 



Total 



556 



1962 



685 



Most of the olive green glass consisted of wine bottle fragments (Figure 77). 
The clear and green tinted glass generally represented small items, mostly 
pharmaceutical bottles, tumblers, or stemmed glasses. 



There was one example of a ground and faceted clear glass bottle recovered 
from Feature F2 at 38BK76 near Structures 76D and 76M, which probably had a 
glass stopper. This may have been used for perfume. Four reconstructable 
olive green wine bottles were recovered from Feature F8 near Structure 76E. 
Based on form, push ups, and pontil marks, two bottles were determined likely 
to be of English manufacture and two of French manufacture (Jones 1971, 
Douglas and Frank 1972). These have already been discussed in context under 
the discussion of the ceramic to glass ratio at Site 38BK76 in Chapter VII. 



251 



The bottle glass was then divided into minimum vessels, primarily on the ba- 
sis of color under the assumption that only rarely would two bottles have 
exactly the same color and, therefore, chemical makeup. Color was determined 
by comparing all sherds from associated proveniences using a light table. 
Under such conditions, even the most opaque olive green glass gives a distinc- 
tive color. However, especially thick olive green glass may produce a darker 
color than thinner sherds although the quality (i.e. chroma) may be recogniz- 
able; for this reason, the following minimum vessel count may not be more 
than 95 percent certain (Table 40). 

TABLE 40. Glass Minimum Vessels 







38BK75 




38BK76 




38BK245 










Green 




Green 




Green 




01 i ve 


Clear 


Tinted 


Olive Clear 


Tinted 


Olive Clear 


Tinted 


Wine bottle 


19 






73 • 




44 




Case bottle 


6 






7 3 




1 




Medicine bottle 




3 


2 


2 


5 






Unidentifiable 
















bottle 


1 


4 


5 


2 7 


1 


4 


4 


Perfume bottle 








1 








Stemmed glass 




10 




11 




4 




Tumbler 




3 




3 








Pressed glass 








2 








Total 


26 


20 


7 


82 29 


6 


45 8 


4 





Sites 38BK76 and 38BK245 contained relatively more olive 
38BK75. This may be a reflection of the increasing use of 
clear glass over time. Further, there was an indication 
38BK245 had proportionally higher amounts of liquor bottles 
ties or drinking glasses than 38BK75. Since this might be 
slave of status overtime, the material was grouped (Table 
the differences between the three major forms. 



green glass than 

green tinted and 

that 38BK76 and 

to medicine bot- 

an indication of 

41) to illustrate 



TABLE 41. Major Glass Forms 
as a Percentage of Glass Artifacts 





38BK75 


38BK76 


38BK245 






# 


% 


# % 


# 


% 




Liquor bottles 
Medicine bottles 
Glasses 


25 

5 

13 


58.14 
11.63 
30.23 


83 79.81 

7 6.73 

14 13.46 


45 

4 


91.84 
8.16 




Total 




43 


104 






49 





252 



This table more clearly shows the difference in relative "popularity" of the 
three functional categories. However, since the percentages reflected popu- 
larity only within glass at each site, liquor bottles appeared to be more 
popular at Site 38BK76 and 38BK245 than at 38BK75. The following chart 
(Table 42) puts the forms into perspective by taking them as a percentage of 
the overall artifact count, using minimum vessel counts. 

TABLE 42. Major Glass Forms 
as a Percentage of Total Artifacts 





38BK75 




38BK76 




38BK245 




# 


% 


# 


% 


# % 


Liquor 

Medicine 

Glasses 


25 

5 

13 


.39 
.08 
.20 


83 

7 
14 


.41 
.03 
.07 


45 .92 
.00 
4 .08 


Total Artifacts 


6689 




22666 




5942 





Table 43 shows that as a proportion of the total number of artifacts, liquor 
bottles were more frequent at 38BK245 than at either of the other two sites; 
Sites 38BK76 and 38BK76 had approximately the same proportion of wine and 
case bottles; and Site 38BK76 had more medicine bottles and glasses than 
either 38BK76 or 38BK245. 

Site 38BK76 was hypothesized to be the most recent site and as such its inhab- 
itants should have been the most acculturated into Anglo-American culture; 
therefore, it should follow that the site had a higher proportion of European 
items such as drinking glasses. This appeared to be the case since 38BK76 
had over twice the amount of drinking glasses than were found at the two 
earlier sites. 

Gun Parts 

Gun parts are discussed here since it has been thought by some archaeologists 
(Fairbanks 1972:84 and Otto 1975:354) that slave access to guns was nonexist- 
ent. As will be shown below, this was not the case at Curriboo and Yaughan. 
A total of 33 arms group artifacts were recovered during the fieldwork (Table 
43), more specimens than appear in drinking glasses, for example. The gun 
parts represented one brass trigger guard from the surface at 38BK76 and a 
rifle or musket barrel from Feature F3 at 38BK245. The gun barrel was 
X-rayed and the inside diameter was determined to be .625 caliber. This was 
too large for the average American Revolutionary period rifle, but was almost 
ideal for the French Charleville rifle (Ferguson 1978:66). 



253 



TABLE 43. Gun Parts and Associated Artifacts 





Gun 


spalls 


Gunf 1 i 


nts 


Shot 


Musketballs 


Gun 


Parts 


Site 
Total 


38BK75 




1 






9 


1 







11 


38BK76 






2 




1 


1 




1 


5 


38BK245 




8 






3 


5 




1 


17 


Total 




9 


2 




13 


7 




2 

























The gunspalls were irregularly shaped gray or gray-brown chert, usually with 
a bulb of percussion and secondary retouch. Gunflints were well made of 
honey colored flint. Their shape and manufacture closely followed that des- 
cribed by Stone (1970:21, type SAT1). 

The lead shot were small, B-B size or smaller buckshot, undoubtedly used in 
shotguns. The musketballs, on the other hand, were large, lead balls, usual- 
ly flattened from impact, used for hunting larger game. Four of the speci- 
mens noted for 38BK245 were given to IAS-Atlanta by construction workers who 
claimed the balls came from the site. These four examples were imbedded in a 
soil matrix in a British Brown stoneware jug which has been partially 
restored. 

The gun parts at Site 38BK75 were mainly found in Features F27 and F29, with 
very few being found during excavation or on the surface. At 38BK76, none of 
the Arms Group artifacts were found in features, and the most spectacular 
artifact, the trigger guard, was found on the surface. At 38BK245, the Arms 
Group artifacts were fairly evenly divided between feature and nonfeature con- 
texts. This may imply that such artifacts played a stronger role at 38BK75 
and 38BK245 where there were enough such artifacts for them to be deliberate- 
ly discarded in features. 

If it is assumed that firearms were used primarily for hunting on the planta- 
tions, one should expect to find evidence of wild fauna being consumed either 
in the slave quarters, or the overseer's or owner's house. There was minimal 
evidence for a wild animal diet at 38BK75 and 38BK76, 1.97 grams of goose 
bone and 6.5 grams of deer bone at 38BK75 and none at 388K76. At 38BK245, 
there were 1.64 grams of opossum bone. All of the remaining bone or shell 
was from freshwater fish, clams, oysters, and domestic mammals. Hunting 
apparently played little part in the slave diet. The conclusion drawn from 
this is that if the presence of firearms reflected hunting by slaves, then 
their catch was probably consumed by the overseers or owners; this is similar 
to the conclusion arrived at by Otto for St. Simons (Otto 1975:287-356). 
Firearms may also have been used by slaves to chase off animal pests in the 
fields. 



254 



Tobacco Pipe Group 

Two major types of tobacco pipes were present at the sites and may have 
reflected slave lifeways. These include "ball clay" pipes, and Colono pipes, 
made of the same clays as Colonoware ceramics. 

The "ball clay" pipes generally exhibited little decoration. At Site 38BK75, 
identifiable stamped initials included an impressed "TD" bowl with "WG" on 
the spur and a raised "S" on another spur. According to Iain Walker 
(1967:76), the "S" was a Dutch mark for the lowest quality pipe and is no ear- 
lier than 1739-40. "TD" pipes were eighteenth to early nineteenth century 
and do not help analysis of the sites beyond confirming their general age. 
Site 38BK76 had four impressed "TD" bowls and two raised "TD" spurs. This 
site also had a raised "WG" on a spur, "MON" or "HON" on a bowl, and a raised 
"RT" on another bowl. The "WG" also appeared at Fort Michilimackinac between 
1715 and 1781 (Peterson 1963). According to Petersen, the "RT" stands for 
Robert Tippet of Bristol, 1680-1740, and this pipe may have represented one 
of the earlier pipe fragments. The "HON" or "MON" bowl may have been a 
portion of a slogan such as Dieu et mon Droit or Honi soit qui mal I y pense , 
implying British manufacture between 1730 and 177D (Noel -Hume 1978: Figure 
97). Site 38BK245 was represented by a "TD" bowl, a "TD" spur, a "WG" spur, 
and a "W" or "V" bowl, perhaps originally a "WG". 

Decoration on pipes included fluting, rouletting around the rim, an incised 
line around the rim, and impressed or stamped leaves on the bowl. Two pipe 
fragments exhibited glaze; these were a brown glazed pipestem end at Site 
38BK76, Feature F8, and a green glazed bowl fragment at 38BK245, Feature F12. 
Noel -Hume (1978:302) notes that this was an uncommon practice in the eight- 
eenth century. 

A more direct link between status and pipes may be the use and reuse of 
pipes. Wear marks on pipe stems caused by teeth or deliberate carving of the 
pipestem occurred only at Site 38BK76. The pipestems were examined for tooth 
marks (sometimes stems were nearly worn in two) and carved stems which would 
have allowed broken stems to take a reed extension, prolonging the life of 
the pipe. Note was also made of the co-occurrence of original or broken stem 
tips. Table 44 was compiled from excavated material. 



TABLE 44. Modified Pipe Tips at Site 38BK76 





Ori 


Worn 
ginal Tips 


Worn 
Broken Ti 


ps 


Carved 
Broken Tips 


Structure 76A 
Structure 76B 
38BK76 




1 
1 


3 

13 
1 




2 
2 


Total 




2 


17 




4 





255 



Few of the toothmarks were on original stems and the vast majority were on 
broken stems. This illustrated long use of pipes and reuse of broken pipes, 
which did not occur at Sites 38BK75 and 38BK245. The implication is that the 
slaves at 38BK76, and especially at Structure 76B, could not afford or were 
not supplied with enough pipes to meet the demand. If this had happened at 
only one structure, it could be considered an idiosyncracy of a single slave; 
that it happened across the site implies that the slaves at 388K76 did not 
have as ready access to pipes as they did at 38BK75 and 38BK245. 

For slave sites, the relative amounts of stems to bowls may also be indica- 
tive of use and discard. It is hypothesized that in the Anglo-American cul- 
ture, a broken pipestem past a certain point would be cause for discard of 
the entire pipe more often than it would be with slaves, and that slaves 
would continue to use the pipe until the stem was only an inch or two long. 
In the archaeological record, this would result in more stem/bowl and bowl 
fragments relative to stems on Anglo-American sites, and more stem fragments 
relative to stem/bowl and bowl fragments on slave sites. This is hypothe- 
sized to be so, since a pipe broken only once and discarded would result in 
fewer and larger stem fragments while a pipe used despite repeated breakage 
would produce more and smaller stem fragments. Subsequent breakage after 
discard would tend to blur this data, though. Since no Anglo-American sites 
were excavated at the plantations, only slave sites could be compared, and 
the hypothesis changed to fit the circumstances. It was therefore hypothe- 
sized that as economic conditions improved at a plantation and pipes became 
increasingly less expensive, a pattern similar to the Anglo-American slave 
process would be evident in the pipe stem proportions, i.e. at Site 38BK76, 
pipe stems would be relatively more common than at Site 38BK245 and Site 
38BK75. Data are presented in Table 45 for examination this hypothesis. 

TABLE 45. Tobacco Pipe Stems and Bowls 
from Excavated Contexts Only 





Stems 


Stem/Bowl s 


Bowls 


Site 
Total 


38BK75 
38BK76 
38BK245 


89 74.17% 
417 88.16% 
192 83.48% 


9 7.50% 
35 7.40% 
23 10.00% 


22 18.33% 
21 - 4.44% 
15 6.52% 


120 
473 
230 


Totals 


698 


67 


58 


823 





256 



If the stem/bowl category is added to bowls, the following results. 
Stems and Stem Bowls Bowls Total 

38BK75 89 74.17% 31 25.83% 120 

38BK76 417 88.16% 56 11.84% 473 

38BK245 192 83.48% 38 16.52% 230 
r= .9993 

.05 p .01 

Totals 558" 125 WT 



There did appear to be a trend from more stems at 38BK76 to fewer at 38BK245 
and 38BK75. A correlation coefficient was derived to test the strength of 
this trend and it was significant. With comparable samples from Anglo- 
American sites, such a correlation may prove useful in establishing status 
and interplay between pipe users and their overall material culture. 

A further indication that slaves may have had more access to ball clay pipes 
during later periods is the presence of colonoware pipe fragments. Site 
38BK76 had eight pipe fragments, and Site 38BK245 had four. These represent- 
ed 1.06 percent and 1.29 percent of the pipes at the respective sites. To 
our knowledge, Colonoware pipe fragments have not been found on strictly 
Anglo-American sites. 

Activities Group 

Clasp knives were the largest single category of metal Activities Group 
artifacts (Figure 81). Four typical examples are illustrated in Figure 82. 
Site 38BK76 had the highest number of clasp knives with 13, and 38BK75 and 
38BK245 had 8 and 10, respectively. Such knives were apparently found useful 
and necessary by the slave inhabitants and may illustrate one of the items 
for which slaves traded goods with their owners. 

The hoes (Figure 83) were analyzed for their functional category using data 
provided by Egloff (1982). The categories were based on the angle of the hoe 
and the length to width ratio. According to Egloff (1980), hoes with an 
angle of 73° or smaller were used for weeding, 77° for hilling or planting, 
and 83° for grubbing or soil preparation. The height/width ratio divided 
hoes into broad or narrow. All of the measurable hoes were broad. Included 
with the hoes was a two-pronged hoe from Site 38BK245, which had an unknown 
special function. There were six measurable examples of weeding hoes and one 
each of hilling and grubbing hoes. The following table gives a breakdown of 
hoes by site. 



257 




CENTIMETERS 



PADLOCK 



BONE HANOLED CLASP KNIFE 




INCHES 



CENTIMETERS 



A - FLAT LEAO 

B -MOLDED LEAO OBJECT 

C - BRASS MACHINE HANDLE 



INCHES 



CENTIMETERS 



CHISEL 





etNTiM«ri»* 



A - SICKLE BLAOE 
B - HORSE BITS 




CHISEL 



FIGURE 81 

Activity Group Artifacts 



258 




245 -LIO-5 



■■MBgggBggSS II 1 ''^V 1 ! 1 ' " wH H I'iH r ' 



245-FI2 - 10 




"-.:• .••^'.■r-.v-'.v.S'i?!:. :'•.>'.■: 



,»tf 



76 - F2 - 






CLASP KNIVES 




245 -LIO-5 



KEY 



76A - U20 - 2 



FIGURE C2 

Assorted Metal Objects 

Identified by X-Ray 



259 




* 

L*". 







INCMCS 
CfNTlMCTEftS 




V&H 











TOP -2 BROAO HOES 

BOTTOM LEFT- FROG GIG 

BOTTOM RIGHT- TWO PRONGED HOE 



FIGURE 83 

Activity Group Artifacts 



260 



CVJ 



09 



O 

m 

CVJ 








■o 






0) 






c 






•f"» 






E 






£ 






CD 






•M 






cu 




-*: 


o 


*T 


(J 




CC 


O 


00 >> 




_J 


CD <T3 


LU 




c cr 


c£ 


-t-> 


•r- 1 


=> 


(/) 


-i<: x 


CD 


0) 


s_ 


mm 


-C 


o >, 


u_ 


(_> 


3 -O 



261 








-a 








a> 








c 








•^» 








E 








_ 








O) 








— 








0) 








o 




un 








CO 




(/I 


>) 




-^ 


Cn 


m 


LU 


u 


c 


cc 


ac 


o 


•1— 


1 


=3 


I— 


-*L 


X 


O 


■o 


s- 




H-! 


(0 


o 


>> 


U_ 


a. 


2 


-Q 





(A 

o 

3 

o 

< 

(A 



262 



Weeding 



TABLE 46. Measurable Hoes 
Hilling Grubbing Two- prong 



Fragment Total 



Average 
Height/ 
Width 1:1.6 



1:1.6 



1:1.6 



Angle 



Site 



73' 



38BK75 


3 


38BK76 


3 


38BK245 


2 



77° 



83* 



Total 



14 



As would be expected at plantations producing cotton and indigo, the weeding 
hoes outnumbered the other hoes at all three sites. Only one sickle blade 
fragment was recovered from Site 38BK245 (Figure 81), probably representing 
rice or grain harvesting activities at the site. 

The evidence for non-farm tools (Figure 81) in the slave quarters was slight. 
Overall such tools accounted for nine artifacts at all three sites or .03 
percent of all artifacts. These tools included two chisels and a hoof knife 
or sickle blade at 38BK245; two wood rasps, one at 38BK76 and one at 38BK245; 
a wood drill bit, a rat-tailed file, and a whetstone at 38BK75; and an axe- 
head at 38BK76. The major activltes represented were carpentry and mainte- 
nance of equipment. 

The only toys found were clay marbles, one at 38BK75 and four at 38BK76. 
These, along with the small pots mentioned in Chapter VIII, were the only 
clear archaeological evidence of children at the sites. It is, of course, 
possible that the marbles represent gaming pieces used by adults for a game 
such as Oh-Wah-Ree (Zaslavsky 1974). 

Fishing gear was present only at Site 38BK76. This included a rather large 
fish hook, four possible fish hooks, and a four-pronged "frog gig". Small 
fish hooks may also have been present, but probably would not have been 
preserved. With the arms group artifacts, there is arti factual evidence for 
both hunting and fishing at Site 38BK76. 



263 



Unidentified iron, lead, and brass/copper, which make up the majority of the 
Activities Group artifacts, may represent other artifact groups. Certain of 
these artifacts were identified during X-ray analysis and have been included 
in their respective group. Many of the iron fragments are undoubtedly broken 
nails, but which ones could not be determined. The lead fragments may rep- 
resent the raw material for bale seals. These fragments are generally parts 
of flat lead sheets. The brass/copper fragments, only four were found, are 
also from flat sheets. Other Activities Group artifacts represented are 
barrel hoops, one from each site and harness tackle including buckles and 
bits, two at 38BK245, one at 38BK75, and four at 38BK76. 

Locks are included in two places in South' s (1977a) pattern. Door locks are 
included with architectural hardware; other locks are included with the 
Activities Group. One chest lock (Appendix E and Figure 84) was recovered at 
Site 38BK245 and one padlock was recovered from Site 38BK75 (Appendix E and 
Figure 85). Both locks were identified from X-ray. The working of the lock- 
ing mechanism of the padlock was reconstructed after X-raying the lock from 
the front and side. This padlock was found in association with Structures 
75B and 75B2 and resembles one illustrated by Noel Hume (1978: Figure 80) 
dating to 1770-1780 or perhaps earlier. 

These artifacts (and to a lesser extent the artifacts discussed in Appendix 
E) and their relationships with one another and between sites present a 
picture of slave life which is only hinted at in published accounts. Most 
published accounts deal with the nineteenth century, in any case. From the 
artifacts alone one can draw the following characterizations of slave life at 
Curriboo and Yaughan plantations in the eighteenth century: 

- The slaves made the major portion of the pottery they used. 

- The proportion of imported goods increased during the later eighteenth 
and early nineteenth centuries. 

- Slave material culture lacked variety and reflected a much lower economic 
status than was enjoyed by whites. 

- Slave contacts with free Indians may have increased after approximately 
1780 with unknown consequences for both cultures. 

- Slave cooking was probably an informal affair conducted around individual 
or family fires, out of doors. 

- Slave diet consisted of a high proportion of liquid or semi -liquid foods 
and very few dishes requiring plates or flatware. 

- Slaves had access to guns, but apparently did not use them to provide 
wild game for themselves. 

- Slaves also used the aquatic resources around them, but may not have 
benefited personally from their catch. 



264 



Pipe smoking probably played a major role in leisure time activities. 
Otherwise, there was very little material evidence of leisure time ac- 
tivities. In fact, the major activities apparent in the slave quarter 
appeared to have been making ceramics, cooking, and eating, although 
other activities such as music (Appendix E), whittling or woodcarving, 
and marble games were hinted at. 

In sum, slave life, as evidenced by the artifacts alone, appeared to have 
been monotonous and on the bottom of the economic scale with very little 
real improvement as time went on. This corresponds closely with the evi- 
dence presented by the architecture in the previous chapter. 



265 



IX. ARTIFACT PATTERNS 

Introduction 

The third anticipated hypothesis stated in the project research design dealt 
with the patterned occurrence of artifact categories within the sites. That 
hypothesis was stated in the following manner: 

Hypothesis 3 . Patterns of artifact disposal on archaeological 
sites are culturally determined and can be discerned through 
careful analysis. Since African slaves were the products of 
radically different backgrounds than British-Americans, the 
pattern of artifact disposal on African slave domestic sites 
should be discernable from that present on British-American 
sites. The difference will be expressed in varying frequencies 
of the artifact categories that the sites share in common, as 
well as the absence of certain categories (Garrow and Wheaton 
1979). 

Hypothesis 3 was construed as a primary research question for this project as 
it provided the mechanism whereby comparisons could be drawn in a quantita- 
tive fashion with other excavated sites. 

The concept of quantification and organization of artifacts into categories 
following a patterning approach is fairly new in historic archaeology. The 
major impetus for quantification studies was provided by Stanley South' s 
Method and Theory in Historical Archeology , published in 1977. Productive 
comparative studies utilizing quantification and the artifact pattern concept 
have been slow to reach print, but do exist (Lewis 1976; Lewis 1977; Hols- 
chlag, Rodeffer and Cann 1978; Drucker and Anthony 1979; Foss, Garrow and 
Hurry 1979; Otto 1975). 

It is not sufficient to simply excavate a site, count the artifacts, and 
arrange them into South' s (1977) artifact categories to achieve the type of 
comparable results needed for intersite comparison. South (1977:89) stated 
the major problem faced in studies of this type in the following manner: 

One of the problems faced when quantification studies are under- 
taken utilizing all artifact classes is obtaining collections of 
excavated data recovered under comparable conditions. 

The methods employed in the excavations at Yaughan and Curriboo have been 
described elsewhere in this report, but a few points need to be restated at 
this time. The methods employed at Curriboo Plantation (Site 38BK245) were 
dictated by the condition of the site at the time of the investigation. The 
slave domestic occupation area had been machine stripped, and had been fur- 
ther damaged by the action of freeze/thaw and rain. The surface artifacts 
recovered from that area came from features or highly disturbed surface con- 
texts. All feature fill was screened through quarter inch mesh screen, and 
numerous soil samples were taken and processed through flotation. The only 
area of Curriboo investigated that had not been subjected to prior distur- 
bances was not a domestic occupation, and reflected usage as a warehouse/pro- 



266 



cessing structure followed by use as a plantation office. A second nondomes- 
tic area, which contained the heavily disturbed remnant of a brick clamp, was 
also explored. 

The investigations at Yaughan on Sites 38BK75 and 38BK76 were conducted ini- 
tially as large block excavations followed by controlled surface collections, 
machine stripping, feature excavation, and minor structural excavation. The 
block excavations were hand excavated and all soil (with the exception of a 
surface root mat in the case of 38BK76) was screened through quarter inch 
mesh. All feature fill was similarly screened, and standard sized soil sam- 
ples were extracted from each provenience for flotation and small artifact 
recovery. The block excavations at Yaughan were placed so as to bracket in- 
dividual structures, and an attempt was made in each case to include enough 
yard area so that representative external debris was recovered. The ration- 
ale behind the block excavations was to recover arti factual and architectural 
data from individual domestic occupations that could be used with a high 
level of confidence to construct artifact patterns reflective of the material 
culture of the inhabitants. The results of the block excavations were viewed 
as the means whereby the validity of the overall site patterns could be de- 
termined, as well as a control against which the data from more restricted 
contexts could be compared. 

The Carolina and Frontier Artifact Patterns 

Prior to discussing the artifact patterns retrieved from Yaughan and Curriboo 
Plantations it is necessary to consider the existing artifact pattern models. 
Stanley South (1977) has proposed two artifact models, the Frontier and Caro- 
lina Patterns, based on the result of a number of site excavations. His 
Frontier Artifact Pattern was postulated on the basis of three sites: Fort 
Ligonier in Pennsylvania; Fort Prince George in South Carolina; and Spal dings 
Store in Florida. Fort Ligonier and Fort Prince George dated from the French 
and Indian War, whereas Spalding's Store dated from 1763 to the present 
(South 1977:143). South characterized the Frontier Artifact Pattern as hav- 
ing a high percentage of Architecture Group artifacts in comparison to the 
Kitchen Group, but close examination of his data indicates that severe inter- 
pretive problems exist that cast doubt on that model. 

The problems that exist with the Frontier Artifact Pattern Model have to do 
with the natures of the sites used, possibly the manner in which the sites 
were excavated, and definitely with the way in which the artifact patterns 
were constructed. Study of the three sites that make up the Frontier Arti- 
fact Pattern Model indicates that two clustered near one end of the observed 
range while the third exhibited significantly different artifact profiles. 
Fort Ligonier, a British occupied French and Indian War fort, exhibited the 
low kitchen/high architecture percentages that South (1977:146-149) states is 
typical of the Frontier Artifact Pattern Model. The problem that exists with 
Fort Ligonier is that South (1977:152) states that the excavations only sam- 
pled the site, but he does not give any indication of the size of the sample 
and how it was drawn. South (1978) has recently admitted that it is possible 
to recover the Frontier Artifact Pattern from excavations inside a structure 
while retrieving the Carolina Artifact Pattern from outside the same struc- 
ture. At the moment it is not possible to determine if that was indeed the 
case at Fort Ligonier, based on South 's statements. 



267 



The nature of the excavation at Fort Prince George was stated. According to 
South ( 1 977a: 1 52) the entire interior of the fort was excavated. The problem 
with the Fort Prince George Pattern is somewhat more basic than the one des- 
cribed for Fort Ligonier. A large percentage of the artifacts recovered from 
Fort Prince George were Cherokee ceramics presumably accumulated during the 
French and Indian War occupation. South initially placed the Cherokee cer- 
amics in the Activities Group, but then observed: 

From the empirical percentage relationship profiles seen here it 
is apparent that the most deviant percentage is that for the 
Acti vi ti es group from Fort Prince George. When the artifact 
classes for this group are examined, the 2583 Cherokee Indian 
sherds ^re the obvious reason for this 26.4 percent figure. The 
presence of this quantity of Cherokee pottery on the site is 
understandable since a major function of Fort Prince George was 
Indian trade (John Combes, personal communication). This being 
the case, it would be unreasonable to build into a frontier model 
this bias for Class 36, so we will remove it. With this single 
adjustment the relationships shown in Table 15 are seen (South 
1977a: 143-145). 

The adjusted pattern for Fort Prince George is illustrated along with the pat- 
terns from Fort Ligonier and Spalding's Lower Store in Table 47. The Chero- 
kee ceramics were thus dismissed, and a pattern reflecting a high percentage 
of Architecture to a low percentage of Kitchen Group artifacts was produced. 

Perhaps the first flaw in South' s approach was in attempting to place the 
Cherokee ceramics within the Activities Group. South (1977a: 97) stated in 
defense of that placement: 

The Colono-Indian Pattern, Class 36, might functionally be in- 
cluded under the Kitchen group, but is kept under Activities due 
to the expected variability of this class of artifact, and its 
role in indicating Indian contact. 

South thus admits mixing criteria for placement of this artifact class, as 
the entire thrust of the remainder of his basic model is predicated on func- 
tion. He is not bothered by the tremendous variability present in Anglo- 
American ceramics when he places that within the Kitchen Group, but in turn 
uses the variability argument to relegate both Indian and, as will be later 
demonstrated, African slave ceramics to non-kitchen functions. Although 
placement of the Indian- and slave-made wares in the Activities Group has a 
great effect on some of the resulting patterns, the effect of totally drop- 
ping the Cherokee ceramics from the Fort Prince George sample is devastating. 
It does little good to quantify artifacts if the investigator is simply going 
to select what evidence he wants to use and discard the rest. It is to 
South's credit that he discussed his rationale for deleting the Cherokee ce- 
ramics from the Fort Prince George sample in sufficient detail that his steps 
could be reversed. 

The basis for South's Frontier Artifact Pattern Model collapses when the 
Cherokee ceramics from Fort Prince George are placed within the site's Kit- 
chen Group. Table 48 presents the Fort Prince George artifacts with the 



268 



TABLE 47. The Frontier Artifact Pattern As Proposed By South 
(1977: Table 16, p. 145) 



Fort Ligonier 
Pennsylvania 
1758-1766 



Fort Prince George 

South Carolina 
ca. 1753-ca. 1769 



Spal dings Lower Store 

Florida 

ca. 1763-Present 



Artifact Group 



Kitchen 


5,566 


25.6% 


1 


,679 


22.7% 


5,789 


34.5% 


Architecture 


12,112 


55.6 


4 


,252 


57.5 


7,222 


43.0 


Furniture 


44 


0.2 




6 


0.1 


51 


0.3 


Arms 


1820 


8.4 




471 


6.4 


227 


1.4 


Clothing 


833 


3.8 




70 


1.0 


51 


0.3 


Personal 


99 


0.4 




9 


0.1 


10 


0.1 


Tobacco Pipes 


411 


1.9 




851 


11.5 


2,343 


14.0 


Activities 


893 


4.1 




50* 


0.7 


1,077 


6.4 



Total 



21778 100. 0% 



77358 100.0% 



16,770 



TUO% 



*South originally included 2,583 Cherokee sherds in this total, but deleted 
them from the sample because of their effect on the Activities Group percen- 
tage. 



269 



TABLE 48. Adjusted Fort Prince George Artifact Pattern 
Compared to the Fort Watson Pattern 



Fort Prince George Fort Watson Mound Summit 

South Carolina South Carolina 



Artifact Group 1753-1769* 1780-1781 



•• 



Kitchen 4262 42.7%*** 627 43.8% 

Architecture 4252 42.6 595 41.6 

Furniture 6 0.1 19 1.3 

Arms 471 4.7 128 8.9 

Clothing 70 0.7 23 1.6 

Personal 9 0.1 2 0.1 

Tobacoo Pipes 851 8.5 18 1.3 

Activities 50 0.5 20 1.4 

WT\ RJO£ 1415 T0U70T 



* South 1977:143-145 
** South 1977:158-159 
***Cherokee ceramics included within the Kitchen Group 



270 



Cherokee ceramics properly placed and presents the Fort Watson artifact 
pattern as published by South (1977a: 158-1 59). The artifact percentages are 
seen to be amazingly similar, varying more than 1.2 percent only in the Arms 
and Tobacco Pipe groups. Fort Watson was utilized during the Revolutionary 
War, and the elevated Arms Group is explained by South (1977a: 159) as reflec- 
tive of the battle that occurred on the site. Both sites were located in 
South Carolina, and in each case the excavations incorporated the area within 
the stockade (Ferguson 1977:45). The variance in the Tobacco Pipe Group is 
hardly surprising, as this seems to be one of the least stable and most sub- 
ject to idiosyncratic behavior of all of the artifact groups. 

The artifact pattern achieved from Spalding's Lower Store assumes new meaning 
if realigned and compared with Fort Prince George and Fort Watson. A total 
of 167 Colono-Indian ceramic sherds were listed with the Activities Group for 
Spalding's Lower Store (South 1977:161). Tables 49 and 50 illustrate the 
pattern originally presented by South and the slight shift in percentages 
achieved by moving the Colono-Indian ceramics from the Activities Group to 
the Kitchen Group. The revised pattern also reflects moving the single stub 
stemmed pipe from the Activities Group to the Tobacco Pipe Group. Again, the 
rationale for that shift is to place artifacts of similar function in the 
same groups. 

The revised Frontier Artifact Pattern is best understood when compared to the 
Public Structure Pattern that has been proposed by Cara Wise (1978). Wise 
utilized three contexts from the Delaware State House in Dover to construct 
this pattern, and the three contexts chosen were believed to reflect the actu- 
al use periods of the site as a public structure. Wise excavated extensive 
areas of the State House yard, and although her excavations did not incorpo- 
rate a ruin (the State House is still standing and functioning as a public 
building) and yard, the percentages she achieved in the contexts used to de- 
fine the Public Structure Pattern are worthy of discussion. Table 51 re- 
flects the three contexts Wise (1978:119-120) used as the basis of the Public 
Structure Pattern. 

A second site used by Wise to formulate the Public Structure Pattern was Toft 
8 excavated at Camden, South Carolina by Kenneth Lewis (1976:116). Lewis's 
excavations at Camden can best be described as extensive testing, and his 
methods are thus not strictly comparable to those employed in the excavations 
of the sites used to build South *s Frontier Artifact pattern, but again (as 
in the case of Wise) his results are worthy of discussion at this point. 

The third site used by Wise to define the Public Structure Pattern was the 
Hepburn-Reonalds House, which was excavated by South (1977:154-158) at Bruns- 
wick Town. South described the Hepburn-Reonalds House as a deviant from the 
Carolina Artifact pattern, and pointed out that it has been utilized as a 
shop and residence (1977:51). 

The primary element that the Delaware State House, Toft 8 at Camden, and the 
Hepburn-Reonalds House had in common was that all three sites served public 
or mercantile functions. The Delaware State House functioned as a court and 
public offices from its initial occupancy. Toft 8 at Camden contained what 
has been described as a "brew house," thus reflecting its mercantile func- 
tion. The Hepburn-Reonalds House functioned as a shop with an attached 



271 



TABLE 49. The Artifact Patterns From Spalding's Lower Store 



Spalding's Lower Store 
[following South 1977:145: 



Spalding's Lower Store 
Revised 



Artifact Group 



Ki tchen 

Architecture 

Furniture 

Arms 

Clothing 

Personal 

Tobacco Pipes 

Activities 



5,789 

7,222 

51 

227 

51 

10 

2,343 

1,077 



34.5% 
43.0 
0.3 
1.4 
0.3 
0.1 
14.0 
6.4 



16,770 



100.0% 



5,956* 


35.5% 


7,222 


43.0 


51 


0.3 


227 


1.4 


51 


0.3 


10 


0.1 


2,344** 


14.0 


909*** 


5.4 



16,770 



100.0% 



* Includes 167 Colono-Indian Sherds 

** Includes 1 Stub Stemmed Pipe 

***Excludes 167 Colono-Indian Sherds and 1 Stub Stemmed Pipe 



TABLE 50. Revised Frontier Artifact Pattern Model 



















Revised 




Fort 


Prince 






Spalding's 


Revi sed 


Frontier 


Arti fact 


George 






Lower 


Store 


Observed 


Artifact 


Group 


(Revised) 


Fort 


Watson* 


(Revised) 


Range 


Pattern 


Ki tchen 


4262 


42.7% 


627 


43.8% 


5,956 


35.5% 


35.5-43.8% 


40.7% 


Architecture 


4252 


42.6 


595 


41.6 


7,222 


43.0 


41.6-43.0 


42.4 


Furniture 


6 


0.1 


19 


1.3 


51 


0.3 


0.1-1.3 


0.6 


Arms 


471 


4.7 


128 


8.9 


227 


1.4 


1.4-8.9 


5.0 


Clothing 


70 


0.7 


23 


1.6 


51 


0.3 


0.3-1.6 


0.9 


Personal 


9 


0.1 


2 


0.1 


10 


0.1 


0.1 


0.1 


Tobacco Pipes 


851 


8.5 


18 


1.3 


2344 


14.0 


1.3-14.0 


7.9 


Activities 


50 


0.5 


20 


1.4 


909 


5.4 


0.5-5.4 


2.4 




9971 




1432 


1U0.0% 


16,770 


100.0% 




100.0% 



272 



TABLE 51. Artifact Patterns From The Delaware State House Excavations 

(Wise 1978:119-120) 





Topsoi I 


Above 








North 


Lower 




Crushed 


Brick 


Crushed 


Brick 


Topsoi 1 


Artifact Group 


1788- 


1807 




1788 


1742 


-1788 


Kitchen 


1142 


51.4% 


519 




50.5% 


380 


48.0% 


Architecture 


982 


44.2 


440 




42.8 


335 


42.3 


Furniture 





0.0 


1 




0.1 


3 


0.4 


Arms 





0.0 


3 




0.3 


4 


0.5 


Clothing 


32 


1.4 


34 




3.3 


36 


4.5 


Personal 


4 


0.2 







0.0 





0.0 


Tobacco Pipes 


40 


1.8 


24 




2.3 


28 


3.5 


Activities 


23 


1.0 


6 




0.6 


6 


0.8 




2223 


100.0% 


1027 




99.9% 


792 


100.0% 



273 



living space, and was therefore also mercantile in function. The artifact 
pattern achieved from Toft 8 and the Hepburn-Reonalds House are presented in 
Table 52. The Hepburn-Reonalds House Pattern has been adjusted in the same 
manner that the previous South sites have been realigned, with the 8 Colono 
sherds from the site moved to the Kitchen Group. 

The "Public Structure Pattern" as proposed by Wise bears close similarity to 
the revised Frontier Artifact Pattern. The similarities are indeed close 
enough that the two proposed patterns should be combined to form a new pat- 
tern. The components of this new pattern are not as dissimilar as they would 
appear at first glance. The Delaware State House, the Hepburn-Reonalds 
House, and Camden Toft 8 were located within town settings, while Fort Prince 
George, Fort Watson, and Spalding's Lower Store occupied rural settings. 
Despite dissimilar settings, four of the sites shared similar functions. The 
Hepburn-Reonalds House served as a shop, Camden Toft 8 was a "brew house" 
Fort Prince George doubled as a Cherokee Trading Post, and Spalding's Lower 
Store was a British Trading Post. The Delaware State House served entirely 
court and office functions, and Fort Watson appears to have been soley a 
military fortification. Loosely interpreted, all six sites served public 
versus single family domestic functions. 

Table 53 depicts the observed ranges of the revised Frontier Artifact Pattern 
and the Public Structures Pattern. Those categories are reflected by the 
terms "urban," which correlates to the Public Structure Pattern, and "rural," 
which correlates with the revised Frontier Artifact Pattern. Perhaps it is 
significant that the extreme low range Kitchen Group occurrence and the 
extreme high range Tobacco Pipe Group occurrence came from the Spalding's 
Lower Store site. As an example, by dropping the Spalding's Lower Store site 
the observed range for the Kitchen Group becomes 42.7 to 52 percent. The 
Tobacco Pipe Group contracts to 1 to 8.5 percent. The adjusted Public 
Interaction Pattern is reflected in Table 54. 

The concept of a Public Interaction Pattern appears to have value as a com- 
parative unit. It is made up of both urban and rural sites that were not 
products of solely domestic functions. The term "Public Interaction" is 
applicable to this pattern in that the sites are products of public access 
and use, such as was found with the Delaware State House and the shops or 
trading posts, or specialized community interaction as in the case of the 
military installations. The term "Public" Pattern would be insufficient to 
describe the manner in which those sites were produced, as that would set the 
pattern apart as incorporating all public-related sites. It is anticipated 
that at least some public related sites (such as most types of industrial 
sites) would not conform to this pattern. At any rate, the Public Interac- 
tion Pattern is sufficiently well described in the preceding paragraphs to 
provide a comparative unit for the two domestic-related patterns that will be 
described in the following sections. 

The second artifact pattern model proposed by South (1977:83-139), the Caro- 
lina Artifact Pattern, was based on seven contexts at five sites. Problems 
exist with the model as stated by South, and a close examination of these 
problems is necessary prior to any attempt to utilize his results. 



274 



TABLE 52. Artifact Patterns From The Hepburn-Reonalds House* 

And Camden Toft 8** 



Artifact Group 



Camden Toft 8 



Hepburn-Reonalds House 



Kitchen 

Architecture 

Furniture 

Arms 

Clothing 

Personal 

Tobacco Pipes 

Activities 



966 
824 

1 



16 
41 



52.0% 

45 
0.0 
0.01 
0.0 
0.0 
1.0 
2.0 



3714 

3953 

18 

12 

24 

4 

374 

84 



45.4% 
48.3 
0.2 
0.1 
0.3 
0.1 
4.6 
1.0 



TS4T 



100. OH 



ST5T 



100.02 



* South 
**Lewi s 



1977:126-127, adjusted with Colono sherds moved to Kitchen 
1976:116 



TABLE 53. The Public Interaction Pattern 



275 



Observed 
Range-Urban 



Observed 
Range-Rural 



Artifact Group 



Combined 
Observed Range 



Kitchen 

Architecture 

Furniture 

Arms 

Clothing 

Personal 

Tobacco 

Activities 



45.4-52.0% 
42.3-48.3 
0.0-0.4 
0.0-0.5 
0.0-4.5 
0.0-0.2 
1.0-4.6 
0.6-2.0 



35.5-43.8% 
41.6-43.0 
0.1-1.3 
1.4-8.9 
0.3-1.6 
0.1 

1.3-14.0 
0.5-5.4 



35.5-52.0% 
41.6-48.3 
0.0=1.3 
0.0-8.9 
0.0-4.5 
0.0-0.2 
1.0-14.0 
0.5-5.4 



276 



TABLE 54. The Public Interaction Pattern 
Adjusted To Exclude Spal dings Lower Store 

Artifact Group Observed Range Mean 

Kitchen 42.7-52.0% 

Architecture 41.6-48.3 

Furniture 0.1-1.3 

Arms 0.0-8.9 

Clothing 0.0-4.5 

Personal 0.0-0.2 

Tobacco 1 .0-8.5 

Activities 0.5-2.0 



277 



Two of the contexts utilized by South were from the Signal Hill Site (Jelks 
1973) which is located in Newfoundland. The remainder of the sites used by 
South are located in the southeastern United States, and at this point ar- 
tifact pattern studies are not sufficiently advanced to determine if region- 
al, national, or worldwide patterns exist. Mixing Signal Hill with south- 
eastern sites to form a basic model thus becomes somewhat suspect. The more 
important reason for dropping Signal Hill 4 and 9 from the Carolina Artifact 
Pattern at this time, though, is that not all of the artifact patterns from 
those contexts are based on actual counts. The nail counts were not included 
in the site artifact lists, and South (1977) estimated the nail counts based 
on relative percentage of occurrences on southeastern sites. South 's esti- 
mates may be entirely correct, but there is no way of determining that at 
this time. It should also be noted that construction types might be one of 
the most variable patterns between a subtropical climate and a northern cli- 
mate. 

Two other contexts used by South (1977) in the Carolina Artifact Pattern 
Model are suspect for other reasons. Fort Moultrie, located on the South 
Carolina coast, was investigated and reported by South (1974). Fort Moultrie 
was a special function site within which occupation occurred. A compelling 
reason for dropping Fort Moultrie from the Carolina Artifact Pattern is the 
fact that it housed a large number of slaves during at least some points in 
its history, and their presence was amply reflected in the high percentage of 
Colono ceramics recovered by South (1977). South placed the Colono ceramics 
within the site's Activities Group. It cannot be determined from available 
research whether or not he keyed on slave occupied areas of the Fort, but 
moving the Colono ceramics from the Activities Group to the Kitchen Group 
would significantly change the results he achieved. At any rate, Fort 
Moultrie A and B should be dropped from the Carolina Artifact Pattern until 
the identity of South' s research population can be resolved. 

Removal of the Signal Hill and Fort Moultrie sites from the Carolina Artifact 
Pattern Model leaves three of South 's original seven contexts for use in a 
Revised Carolina Artifact Pattern Model. The three remaining contexts (at 
three sites) do require additional adjustments before they can be employed in 
the Revised Model, however. There is little doubt, based on the Cooper River 
Historic Sites Investigations and the results achieved by Drucker and Anthony 
(1979) at Spiers Landing, that Colono ceramics should be placed within the 
Kitchen Group. The sites that compose the suggested Revised Carolina Arti- 
fact Pattern Model listed in Tables 55 and 56 include Colono ceramics in the 
Kitchen Group where they properly belong. 

The Carolina Slave Artifact Pattern 

The results of the historical research indicate that Yaughan and Curriboo 
Plantations were in the early stages of development in 1745. Isaac Cordes 
owned an interest in both plantations at that time, and his estate inventory 
placed livestock and tools on those plantations in that year. Study of the 
historical records and the results of the artifact analysis indicate that 
Site 38BK76 represented the earliest slave quarter at Yaughan Plantation, and 
that that quarter was almost completely abandoned in favor of the slave quar- 
ter at 38BK75 by around 1795. Site 38BK75 was apparently occupied by around 



278 



TABLE 55. Revised Carolina Artifact Pattern 
Observed Range and Mean 

Revised Carolina Artifact Carolina Artifact 

Artifact Group Pattern-Observed Range Pattern Mean 

Kitchen 51.8-67.0% 59.5% 

Architecture 25.2-31.4 27.6 

Furniture 0.2-0.6 0.4 

Arms 0.1-0.3 0.2 

Clothing 0.6-5.4 3.0 

Personal 0.2-0.5 0.3 

Tobacco Pipes 1.8-14.0 7.8 

Activities 1.0-1.9 1.3 



279 



Cn — 

O 

CO CM 

CT>00 

■O .-I 
t- I 
1- CO 
j3 CO 
1= P» 
(O «— ! 

lo »— 



a. 
o 



< 



o 

o 

■o 
(/I 

> 
an 
<u 



<0 

in 



en 

< 



oo ^-« 

O 

^ cn 

O 00 

i— «-4 

2 I 

</i 00 

c CM 

ICQ — - 



LT> 
C\J 
!/> — • 

'0 

r«. 
o r-« 
f— t—t 

2 I 

</> CM 

c co 

3 C-» 
J- -H 

lea — 



Q. 
3 
O 
i- 
C3 

-t~> 
u 
<o 

<*- 

•^ 
■M 

s- 
< 



a* 
cn 


8* 

CO 

1— 1 


a« 
CO 

i— ( 


a« 


a* 

CO 
co 


a* 

U"> 


a« 

'0 

px 


lO 
CO 


o 




in 

CM 






10 




i— I 


I— 1 


o 
o 



10 


(0 


m 


r» 


cn 


CO 


en 


o 


o 


<— 1 


o 


CO 


CM 


<0 


o 


^i- 


p- 


CO 


en 


o 






O 


f* 


CO 


CO 


CO 


CM 


in 






t— i 








en 



a* 


a* 


a* 


a* 


a* 


a« 


a* 


o 


CO 


CO 


«* 


in 


in 


Ti- 


CO 


CO 


lO 


CO 


in 


«— i 


en 



in 



CO 



a« 



co —■ 



a* 
o 
o 

o 
o 



in 


io 


CM 


in 


CM 


O 


cn 


cn 


CO 


en 


»-H 


CO 


•a- 


p*. 


CM 


CM 


m 


«— 1 


co- 


r-H 










CO 


#— 


«— « 


in 


•* 










1—1 




CO 



a« 



io 



a« 

p— 



CM 



a* 
co 

CM 



a« 
cn 

o 



a* 



a* 
cn 



cn ^ 

CM 



a* 

o 



as 
cn 



a» 
o 

o 

© 

o 



o 


O 


CO 


«* 


o 


*H 


o 


p*. 


in 


f— 1 


CM 


CO 


CO 


px 


px 


CO 


*t 


m 


r*. 


IO 






o 




00 


CO 


r*. 


CM 


cn 






.-H 




CM 




10 


CM 
















CO 



o 

r- 
i— < 



on 
•=> 
^— 
o 



o 

or 

< 



Ui 



C3 



00 







* 








* 








* 








oo 








UJ 




_l 




•— « 




< 




1— 




z 




■— 1 


«/> 


o 


oo 


r> 


(— 


oo 


UJ 


1—4 


<o 


or 


o_ 


t— 


+* 


U] 


1-^ 


o 


o 


o. 


Q. 


< 


I— 



cn 
co 

«— « 
i 

CO 

CO 

px . 
p*. </> • 
cn o °o 

J= l_ r— 

4-> CU <U 

3 <J -O 

O 

OO O W 

c u 
E Oi- 

O r- E 

V. O <o 

««- C_) t_ 

cu 

•a io o 

(U cu 

•^ -o o 

<«- 3 C 

•»- i— o 
-o O <— 

o c o 
s: >-i o 

* * * 



280 



1784, when the owner of the plantation, Thomas Cordes, married and began to 
expand his operation. Thomas Cordes apparently doubled the number of slaves 
at Yaughan from the mid-1780s to the mid-1790s, when financial reversals 
forced him to reduce his si avehol dings back to around 40. It is our assump- 
tion that Site 388K75 dates from the expansion period, and that 38BK76 was 
abandoned in the 1790s as Thomas Cordes began to reduce his holdings. This 
assumption is borne out by ceramic bracket analysis (South 1977:219-220) 
conducted on the datable ceramics from the two sites, but is not substanti- 
ated by the Mean Ceramic Dates (South 1977:220) from the two areas. 

The Mean Ceramic Dates (Figure 86) on Sites 38BK76 and 38BK75 appear to be 
contradictory and confusing on the surface. The overall MCD of 1773.0 on 
Site 38BK76 was 3.01 years later than the proposed 1770 mean occupation date 
for the site. The proposed occupation range on Site 38BK75 of 1784-1826 was 
based on historical documentation and ceramic bracket dating, providing a 
mean occupation date of 1805. The MCD on Site 38BK75 was 1789.8, or 15.24 
years earlier than the proposed mean occupation date. The early MCD on 
38BK75 was not surprising in view of the results Drucker and Anthony (1979) 
achieved at Spiers Landing, which was another slave occupied site in Berkeley 
County, South Carolina that dated from the same period as 38BK75. In fact, 
time lag (Adams and Gaw 1977) is to be expected on sites occupied by Afro- 
American slaves (Braley 1980). The reverse of time lag which occurred on 
Site 38BK76 does require explanation, however. 



The concept of mean ceramic dating is based on the premise that datable ceram- 
ics will be acquired, broken, and discarded at a fairly uniform rate through- 
out the occupation history of a site. If a degree of uniformity of acquisi- 
tion and discard occurs, then application of the regression formula that 
forms the heart of the mean ceramic dating concept should produce a date that 
is roughly equivalent to the mean occupation date of the site. The major var- 
iables that can materially alter the results of the MCD if acquisition and 
deposition are uniform and the ceramics are correctly identified would then 
be use of erroneous dates for the individual ceramic types or excavation sam- 
ple error. We feel that we can rule out significant ceramic identification 
errors and excavation sample error in the case of Site 38BK76. Also, the 
manufacture dates utilized are date ranges that have produced accurate MCDs 
on numerous other eighteenth century sites (South 1977). If those variables 
have been successfully controlled, then the most likely answer for the late 
MCD at Site 38BK76 relates to the process of ceramic acquisition and discard 
that took place on the site. 

The ceramic collection from Site 38BK76 was composed of 88.9 percent Colono 
with 11.1 percent imported types. The ceramic collection from Site 38BK75 
consisted of 64.2 percent Colono and 35.8 percent imported types. Several 
minor differences existed between the Colono from Sites 38BK75 and 38BK76, 
but the major variance was the fairly high percentage of jar forms at 38BK76, 
and their virtual absence at 38BK75. The form shift in itself does not nec- 
essarily account for the relative percentage of Colono to imported ceramics 
on the two sites because the jar forms appear to have been primarily used as 
cooking vessels, and that functional niche was apparently not filled by im- 
ported vessel types. The interpretive key in this case probably relates to 
the economic and social impact that the Revolutionary War had on the general 



281 



<0 
CM 

CO 



a :*£ 

■M CQ 

l/> <T3 CO 

(U Q CO 



a 



WO 

00 





C T3 

O C 

•.- (^ 

ro WO 
Q-l^ 

3 ^ 

U CO 

u co 

o en 

C CD 

ro •*-> 

OJ •,- 
</1 



=5 

•— ^r c o 






282 



area. That is, it appeared that more goods from outside the immediate envir- 
ons of Yaughan Plantation became available to the slaves during and/or after 
the Revolutionary War as the virtual economic/social isolation of the slaves 
was broken. This, of course, assumes that the Colono was being made at 
Yaughan by slaves for their own use, which we feel we have established on the 
basis of our research. If these interpretations are correct, then it is 
reasonable to assume that a much smaller percentage of imported wares was 
utilized at Yaughan prior to the Revolutionary War, and that the ceramic 
assemblage on the site underwent a dramatic shift during and/or after the 
war. The late MCD on Site 38BK76 then becomes a function of the irregular 
acquisition and discard patterns that were operative on the site. In this 
case the MCD ceases to be a dating tool and becomes a key to understanding 
the nature of the Site 38BK76 ceramic assemblage through time. 

If the interpretation of the cause of reverse time lag on 38BK76 is correct, 
then the presence of time lag on 38BK75 indicates that the ratio of Colono to 
imported ceramics probably remained fairly stable through the occupation his- 
tory of Site 38BK75. That interpretation can be very important to future re- 
search on Afro-American slave-occupied sites in the area because it would 
seem to indicate that there was not a gradual decline in the utilization of 
Colono wares through time as has been generally assumed. It seems more like- 
ly based on our research that the eventual abandonment of Colono wares will 
be found to have proceeded as a few radical shifts rather than as a gradual 
decline. 

Understanding the nature of the ceramic assemblages at Site 38BK76 and 38BK75 
through time enhances understanding of the artifact patterns extracted from 
the sites. As an example, if it is correct to assume that a majority of the 
imported ceramic sample at Site 38BK76 dated from late in the occupancy of 
the site, then it would follow that at least some of the other nonlocal ar- 
tifact categories followed a similar acquisition/discard pattern. The arti- 
fact patterns from Sites 38BK75 and 38BK76 (Table 57) are surprising enough 
as they stand, but they take on an even greater meaning if it is assumed that 
the 38Bk76 pattern was buffered in the manner previously described. 

The artifact patterns extracted from Yaughan Plantation, when combined with 
the results from Curriboo and Spiers Landing (Drucker and Anthony 1979) are 
distinctly different from the Revised Carolina Artifact Pattern and the 
proposed Public Interaction Pattern. The Public Interaction Pattern, with 
its nearly equal Architecture Group to Kitchen Group counts, is easily dis- 
tinguished from what we propose to call the Carolina Slave Artifact Pattern. 
The Revised Carolina Artifact Pattern, bears closer resemblance to the per- 
centages that characterize the Carolina Slave Artifact Pattern, but becomes 
less similar if modified in response to what has been learned from the in- 
vestigation of the Cooper River Historic Sites. 



283 



o_ 
u 



4-> 

< 

> 
to 

Lo 

<o 

c 



o 

i- 

0) 



CO 

< 



H 1 




■K 




* 




CD 




e 


^^ 


>i— 


o 


■o 


oo 


c 


00 


<T3 


t— 1 


_J 


1 




o 


</> 


en 


S- 


r-. 


O 


.— i 


■»— 


-^■^ 


a. 




to 




LD 




r^ 


^■■^ 


AS 


tO 


GQ 


CM 


ao oo 


ro 


«— t 


c 


^> 


fO 00 


J= 


*>* 


CT)«— 1 


3 


*«*» 


<T3 




>- 




t 




LO 




**• 




CM 




n 


^^ 


ea 


o 


CO o 


no oo 




1—1 


o 


1 


o to 


j=> 


«3" 


»f— 


r^ 


u 


^ 


s- 


- - 


3 




O 




X> 




^ 




J* 


^^ 


CQ 


to 


CO 


en 


oo 


r»» 




<— t 


c 


i 


A3 


to 


JZ 


*r 


cnr^ 


3 


t-H 


*3 


^_ . 


>- 




a. 




3 




O 




%m 




as 




■M 




O 




UJ 




4- 




^— 




+-> 




(. 




< 





CO 



v© r** 
r^ o 



o 

CVJ 



en 

r-. 



»* 9« 8* 

r^» oo to 
O * CO 






o 

CM 



CVJ 



lO 


^H 


CM 


to 


**■ 


CM 


Ti- 


to 


o 


r-« 


OO 






CM 




ro. 


CM 


Tf 


CM 


to 














o 


CVJ 
















oo 



<y*« tT^ 9^ O 1 Q~ 

0O O CO CO «— • 

r» o o r-* to 



8« 


S* 


»« 


VO 


o 


*!■ 


O 


en 


io 



o 



to 

CM 



en en 
oo vo 

*r to 



to ~* 



CM 

oo 



CM 



CM 
CO 



oo 



p— 

CM 
VO 



a* »* W 

p«- vo r*» 
r» to o 



CM 



to 

OO 



*T »-H CM 

o **• *j- 



en 



O 

CM 



0O 



to 



IO 



to o 

-4 CM 



CM 



OO 



OO 


<— I 


CM 


*»■ 




to 




to 



o 

CM 



JtojO wA "k^D ->- C\ ^O >i P 

lj^ vT^ vT^ v»^ tT^ ^** 

cm to cm o oo r*. 

00 O O OO O OO 



r— I 

CM 



CO -* 



oo 



o 


o 


CM 


to 


VO 


to 


CM 


VO 


r- 


o 


Tj- 


t— 1 




to 




to 


«* 


CM 


00 


to 










r»- 




OO 


00 


CM 














CM 


«-H 
















CM 

















* 


















* 


















* 






UJ 












* 






or 












* 






3 












CO 






t— 


UJ 








* 


UJ 






O 


or 




to 


_i 


* 


M 




z 


UJ 


3 




2 


«c 


* 


1— 




UJ 


t— 


t— 




■— ) 


2 


* 


NM 


</» 


■jz 


•— < 


H^ 




:r 


O 


tst 


> 


r— 


o 


3: 


z 


oo 


>— 


00 


UJ 


^4 


<o 


>— 


CJ 


or 


X 


o 


or 


o. 


»— 


4-> 


1—4 


cc 


=> 


or 


_j 


UJ 


1-^ 


o 


o 


^ 


< 


U- 


< 


<_> 


o. 


CL 


< 


I— 



en 
en 



CL 
3 O 

O to 
s- *r 
ts CM 



>» 

c 
o 



c a» 
<u 

J= 3 

♦^ <J 

t- 3 

^ S- 

4-> 

C UO 

(U 

^■§ 

"O r- 
3 O 

r— C 
<J •*- 

C 



O </) 
i— <u 

o o 



t. -a 



0) 



&. tn <u 
a> <u "a 

^ o. 

3 a.'io 

o o a> 
- c s 

O i— "O 

s- o cy 

«4- <_) ^~ 

<4- 

•O irt -^ 

CU 0) 4-> 
f- "3 C 
<♦- 3 <U 
•»- r- "3 
-3 <J t- 

O C C 

Z •"-" => 

« * -x 

* * * 

* * * 



284 



In summary, South 's Carolina Artifact Pattern was based on the results of the 
excavation and analysis of seven sites. Two of those sites, denoted Signal 
Hill 4 and Signal Hill 9 (Jelks 1973), should be struck from the Carolina 
Artifact Pattern for two reasons. The first reason is that the sites are 
located well outside the southeastern United States (Newfoundland), and may 
not be reflective of the patterns of acquisition and discard operative on 
distinctly southern sites. The second, and most important, reason to drop 
those sites is that the nail percentages within the Architecture Group are 
not based on actual artifact counts, but instead represent estimates based on 
artifact ratios from southern sites. Two more sites, both within Fort Moul- 
trie (South 1974, 1977) on the South Carolina Coast, should also be dropped 
from the Carolina Artifact Pattern. Fort Moultrie was a special function 
site that should not be used for comparison with purely domestic sites. 
Also, Fort Moultrie housed a large number of slaves at at least some points 
of its history, and their presence was amply reflected by the high percentage 
of Colono ceramics that South placed within the site's activities groups. It 
is not known on the basis of South' s excavation and research whether or not 
he keyed on slave occupied areas of the Fort, but revision of his Fort Moul- 
trie A and Fort Moultrie B patterns results in artifact patterns that would 
comfortably fit within the proposed Carolina Slave Artifact Pattern. We pro- 
pose simply dropping Fort Moultrie A and B from the Carolina Artifact Pattern 
for the moment until the question of the identity of South' s research popula- 
tion can be resolved. 

Removal of the Signal Hill and Fort Moultrie sites from the Carolina Artifact 
Pattern (CAP) leaves three sites of South 's original seven for comparison 
with the proposed Carolina Slave Artifact Pattern (CSAP). The three remain- 
ing sites require an additional adjustment before the CAP and CSAP can be 
compared, however. There is little doubt based on our work and the results 
achieved by Drucker and Anthony (1979) that Colono ceramics should be placed 
within the Kitchen Group. The sites that now compose the Revised Carolina 
Artifact Pattern shown in Table 58 reflect that revision, with the Colono 
ceramics moved from the Activities Group to the Kitchen Group. 

The Carolina Slave Artifact Pattern can be distinguished from the Revised 
Carolina Artifact Pattern on a number of points (Table 58). The most obvious 
differences are the elevated Kitchen Group and lowered Architecture Group 
percentages within the Carolina Slave Artifact Pattern. Perhaps the most 
important comparative factor resides within the relationship of the joint 
Kitchen-Architecture Groups found in the two patterns. The vast majority of 
the durable material culture from the four slave occupied sites of the CSAP 
falls into the Kitchen and Architecture (subsistence-shelter) Groups. The 
combined Kitchen-Architecture Group artifacts account for 96.02 percent of 
the recovered artifacts at 38BK76, 93.43 percent at 38BK245, 95.73 percent at 
38BK75, and 95.60 percent at Spiers Landing. In comparison, within the 
Revised Carolina Artifact Pattern, Brunswick S25 has a combined Kitchen- 
Architecture of 87.94 percent, Brunswick S10 totals 83.18 percent, while 
Cambridge 96 reflects a total of 90.15 percent. The combined Kitchen-Archi- 
tecture on sites within the CSAP thus averages 95.20 percent of the total 
recovered assemblage. The average on sites within the Carolina Artifact 
Pattern is a somewhat lower 87.09 percent. If pipes are added to these 
figures, an average of 98.73 percent of the recovered artifacts on the 
slave-occupied sites fall into the three categories, leaving a very sparse 



285 



c 



<u 



a. 






4-» « 

i_ CL 

< 

C <0 



o -m 



<« 


< 


o 






at 


4» 


> 


jt 


<o 


-(-> 


p— 




00 


>♦- 




o 


<a 




e 


c 


•^ 


o 


i~ 


^ 


o 


«*• 


u 


s. 


<o 


<T3 


o 


CL 




E 


<l» 


O 


.c 


o 


+-> 


• 




co 

LO 




<u 




-O 




TJ 





0) <u 
> -M 

i— <o 
oo a. 

c u 

r- *+- 

o •«- 

CJ < 



1m 
<U <D 4) 

> +J o> 

<Q 4-> C 
i— IO « 
00 a. Of 

(O 4-> "O 
CUV 

(*- <o > 

— <*- i- 

O >r- a» 
S- +> </> 

LJ < O 



•O C 

C i- 

•»- 0) 

O •♦■> 

<n a. 
o 

"O o 

to <*. 

•^ «f^ 

> ■•-> 

a> t- 

OT < 



o +-> c 

<o to 

<tj a. ar 

■o <_) <y 

<D <T3 > 

</> <4- k 

t- •*- a» 

lor < o 



o 
u 
o 



u 
<a 






9« 
CO 



o 






35 



a« 


a* 


LO 


CO 


o 


LO 



CO 



8* 

CM 

o 

• 

o 
o 



a* 


»« 














o 


o 










** 




CVJ 


o 


a* 


a« 


a* 


a* 


»-« 


8« 


• 


• 


CO 


r«. 


a> 


p-. 


^- 


LO 


**" 


LO 


o 


CM 


r^ 


O 


• 


00 


CO 


CM 


• 


• 


• 


• 


LO 


• 


a« 


a« 


a« 


8* 


a* 


a« 


♦« 


*« 


CO 


CVJ 


LO 


CM 


o 


CO 


CO 


^H 


r*. 


CO 


o 


O 


CO 


o 


^- 


CVJ 


O 


^* 










CM 




r*. 


«— t 















a* 

LO 


a« 

00 
LO 


8* 
LO 
CO 


8* 
CT> 


a* 
LO 


8* 
CM 


a* 

o 

CO 


a« 

CO 


LO 


p*. 

CM 






CM 




r-» 


fi 



O 

o 
o 



a* 


»• 










a< 




r-. 


CO 






a* 




^* 


a« 


a* 


CO 


»« 


«« 


00 


*« 


CT> 


vO 


• 


• 


CO 


^r 


CO 


«3" 


• 


00 


# 


•-H 


vO 


CO 


• 


LO 


CO 


• 


VO 


CO 


• 


• 


LO 


• 


—* 


H 


a* 


a* 


s« 


»« 


a* 


9« 


9 s * 


!»« 


o 


CO 


00 


CP> 


LO 


LO 


VO 


^* 


CO 


.— i 


.— t 


o 


LO 


r-H 


r* 


&l 


«— » 


LO 










«— « 




LO 


CM 















UJ 



O 3C 

r- O 

« or 

^ < 



00 



<>0 



C3 

z 

p— I 

t— 
o 

_J 
o 



o 
or 



oo 



o_ «c 



4-> 

o 



286 



1.27 percent to be scattered over the remaining groups. Combining pipes in 
the Carolina Artifact Pattern figure, there remains 6.11 percent for the 
other artifact groups. 

The Carolina Slave Artifact Pattern is admittedly based on a small number of 
sites, but even so it may reflect some trends worthy of mention. The highest 
Kitchen Group and lowest Architecture Group percentages are found on the 
oldest site (38BK76). The architectural shift that apparently took place 
between 38BK76 and 38BK75 is discussed elsewhere, but it is felt that the 
change in the Kitchen and Architecture Groups from the older to the younger 
site reflects culture change. Indeed, the figures presented in Table 58 may 
represent quantitative acculturation that occurred from the earlier to later 
sites. It is entirely feasible that later nineteenth century sites occupied 
by Afro-Americans will prove to be practically indistinguishable from white- 
occupied sites based on artifact pattern studies. In fact, that should prove 
to be the case if the Carolina Slave Artifact Pattern truly has validity. 



287 



X. SLAVE SUBSISTENCE 
Introduction 

The botanical material was studied by Paul Gardner of the University of North 
Carolina. His report is reproduced with all tables in Appendix G. The faun- 
al material was analyzed by Kay Wood and Elizabeth Reitz of the Zooarchaeolo- 
gy Laboratory, University of Georgia. Their report is included as Appendix 
H. Further analysis of the fauna! material was subsequently conducted by SSI 
in Marietta, and the following discussion includes data from both analyses. 

As noted in the discussions of the features, soil conditions played a large 
role in determining which categories of floral and especially faunal remains 
would be preserved and to what degree. Perhaps the most important condition 
affecting preservation was the existing acidity of the soils at the sites. 
More detailed soils data, including pH, total organic nitrogen, carbon and 
phosphate, for individual structures and features is presented in Appendix C. 
When all pH readings are considered, the following values are obtained: 

TABLE 59. Soil Acidity 

Hand-excavated 
Site Units Features Average 

38BK75 5.99 5.56 5.78 

38BK76 5.96 6.37 6.17 

38BK245 3.97 3.59 3.78 



There was no statistically significant difference between units and features 
within each site. The average readings seriated the sites from least acid, 
Site 38BK76, to most acid, Site 38BK245; Site 38BK75 is very close to 38BK76. 
As pointed out by Wood and Reitz (Appendix H), the natural pH level of mam- 
malian bone tends to be from 7.0 to 8.0. A pH of 3.5 or even 6.0 would 
militate against good bone preservation. This may be even more critical for 
fish bone, since it may be more alkaline than mammalian bone. 

Coupled with disturbance from agriculture at 38BK75 and 38BK245, logging at 
38BK76, and mechanical stripping conducted prior to mitigation at 38BK245, 
the pH levels at Yaughan and Curriboo did not hold out much hope for conclu- 
sive data for faunal analysis. Be this as it may, certain general conclu- 
sions and comparisons could be made which illuminate slave diet and subsis- 
tence patterns with respect to botanical and faunal resources. 

Ethnobotanical Remains 

Obviously, not all plant materials utilized by a site's inhabitants are equal- 
ly likely to be preserved by carbonization. Those plant parts deliberately 
added to the fire as fuel ^re the most likely to be preserved. At the 



288 



opposite extreme, plants used exclusively in areas distant from fires will 
usually be absent from the archaeological record. The proximity to fire 
which a plant is utilized is not the only factor which affects the probabil- 
ity of a plant being recovered archaeologically. Dense plant structures such 
as seeds tend to be preserved in a recognizable form. Succulent plant parts 
such as leaves or tubers are much less likely to be preserved. Hence, evi- 
dence of the utilization of plants for their leaves or roots is difficult to 
acquire. Finally, the way in which a plant is processed may greatly affect 
the probability of its being preserved. The parching of seeds over an open 
fire is highly likely to result in some seeds entering the fire where they 
may be carbonized fairly intact. On the other hand, the grinding of seeds to 
produce flour or meal decreases the possibility of correctly identifying the 
seeds. 

It was decided to focus the analysis on the larger features, since the post- 
holes tended to have little material and included large proportions of non- 
organic material. Unfortunately, it was not possible to analyze all of this 
material. Instead, for the particularly large features, samples of one-half 
to one-fourth of the feature were analyzed. In this way it was possible to 
analyze at least a portion of the material from each clay ex traction/ trash 
pit from the three sites. Although a greater quantity of analyzed material 
and a broader range of proveniences would, of course, be desirable, it was 
felt that the results obtained by the above sampling procedures were suffi- 
cient to characterize the utilization of plant resources at the Yaughan and 
Curriboo Plantations. 

During the course of analysis, 726 grams of material were examined from 61 
provenience units (Tables 60 and 61). Overall, the botanical samples were 
rather trashy, containing a total of 135 grams of material other than car- 
bonized plant remains. The carbonized material included 566 grams of wood 
and pitch, 2.5 grams of maize cupules, 0.8 gram of walnut shell, 1.0 gram of 
hickory nutshell, 10.5 grams of peach pits, 1.5 grams of small seeds, and 7.8 
grams of unidentified fragments, a category which included a mixture of ra- 
ther amorphous pieces, some of which were probably galls, fungus, or bark. 



289 



Table 60. Seeds by Site 
38BK75 38BK76 38BK245 Total 



ft ice 


4 


1U 


5 


19 


Maize 


1 


1 


4 


6 


Peach 


10 


1 


1 


12 


Hawthorn 


1 






1 


Bramble 






1 


1 


Sumac 


1 






1 


Legumes 


6 


7 


1 


14 


Goosegrass 


3 






3 


Unidentified 










Grass: I 


12 


44 


11 


67 


Other Grasses 


1 


1 


3 


5 


Rumex 






1 


1 


Polygonum 






1 


1 


Acalypha 




1 




1 


Unidentified 


41 


49 


47 


137 


TOTAL 


80 


114 


75 


269 





Table 61. Botanical Material by Weight (in grams) 
38BK75 38BK76 38BK245 Total 



Trash 


46.98 


42.87 


135.39 


225.24 


Unidentifiable 










Fragments 


2.02 


2.31 


7.76 


12.09 


Wood and Pitch 


216.56 


167.26 


565.80 


949.62 


Maize Cupules 


.34 


1.38 


2.51 


4.23 


Walnut Shell 


.48 


.25 


.80 


1.53 


Hickory Shell 


.72 


.21 


.98 


1.91 


Peach Pits 


9.83 


.49 


10.50 


20.82 


Small Seeds 


.79 


.42 


1.46 


2.67 


TOTAL 


277.73 


215.19 


725.76 


1218.68 





290 



The majority of the carbonized plant remains recovered from Yaughan (38BK75 
and 38BK76) and Curriboo (38BK245) represented plants used as fuel. This 
included the maize cupules and the wood and pitch. "Pitch" is used here as a 
generic term for any resinous substance exuded by wood as it burns and does 
not refer to a deliberately manufactured naval store. Wood occurred in all 
of the features analyzed, and although no rigorous attempt at species identi- 
fication was undertaken, it can be said with confidence that the overwhelming 
majority of wood fragments were pine, with hardwood fragments being extremely 
rare. Maize cupules were present in 25 of the 51 flotation samples and in- 
dicated the use of corncobs as fuel. 

The presence of maize cupules also strongly suggested the use of maize as a 
foodstuff; this was further indicated by the occurrence of maize kernels in 
four of the samples. This was a surprisingly low number, considering the 
well-established role of maize in prehistoric and historic period diets in 
the southern United States. Of course, the possibility that maize was not an 
important dietary item of the Curriboo and Yaughan slaves could not be ruled 
out absolutely, but the low number of kernels, the large number of cupules in 
the samples, and the regularity with which maize was mentioned as a staple 
food of slaves from mid-nineteenth century South Carolina plantations (c.f. 
Rawick 1972:14, 26, 39, 52, 55, 62, 99, 119 and other passages) seemed con- 
tradictory. The low number of maize kernels may have been the result of 
highly effective milling, which has militated against preservation. Rice 
( Oryza sativa L.) , 19 grams of which occurred in 12 samples, appeared to have 
been an important foodstuff. The one other cultigen which was definitely 
identified was the peach ( Prunus persica ), which, like rice, is a native of 
Asia. Twelve peach pits from nine samples were recovered, but this probably 
did not accurately reflect its true dietary significance. Rather than riv- 
aling rice or maize as a foodstuff as its frequency of occurrence might sug- 
gest, peaches were likely little more than a dietary complement available 
only during a limited harvest season of June to July (c.f. Schopmeyer 
1974:664). Its relative abundance in archaeological sites was largely due to 
the density of the pit, which makes it durable, and to its large size which 
makes it noticeable to excavators. 

One other plant remain may have been derived from a cultigen. Feature F65 at 
38BK245 contained a carbonized plant part, roughly discoidal with a diameter 
of 18 millimeters and a thickness of 9 millimeters. This most closely 
resembled a section of the peduncle (fruit stalk) of one of the Cucurbita- 
ceae, but this identification was far from certain. 

The walnut and hickory nutshells seemed to represent occasional foods rather 
than dietary staples. Walnut occurred in only seven samples and hickory in 
only ten out of the total 61. Furthermore, they were not a large component 
of any flotation sample; in fact, their total combined weight of 1.78 grams 
comprised only 0.3 percent of the total carbonized plant remains recovered. 
This seemed to indicate a very limited exploitation of nuts by the site 
inhabitants. 



291 



The hawthorn ( Crataegus sp.) and bramble ( Rubus sp.) seeds also seemed to 
represent dietary complements, as both were represented by only one seed 
apiece. The fruit of the hawthorn is a small pome with large seeds and a 
small amount of pulp (Fernald and Kinsey 1958). It is, therefore, a food of 
limited appeal, although John Lawson in 1709 described the haws of North Ca- 
rolina as having "... a very pleasant agreeable Taste" (Lawson 1967:112). 
Bramble (a general term for the genus which includes blackberries, raspber- 
ries, and dewberries) can be quite abundant in localized areas and can pro- 
duce a profusion of fruits during its midsummer fruiting season. The sole 
seed of Sumac ( Rhus sp) may have represented a dietary item, as the seeds can 
be used to produce a pleasantly acidic beverage (Fernald and Kinsey 1958). 
It is equally possible, however, that the seed may have derived from a nearby 
plant that had colonized the disturbed habitats created by human activities. 

The other plants which were identified from the samples were unlikely to have 
been of economic importance, and were weedy species which thrive in disturbed 
habitats such as those surrounding human habitations. The 13 legume seeds 
seemed to fit Into this category rather than being domesticated beans. A 
legume from Feature F2 at 38BK75 may have been rattlebox (Crotalapia sp.), 
but it was too distorted to be confidently identified. Five 1 egumes from 
Feature F82 at 388K76 may possibly have been Strophostyles sp., but this 
identification was far from certain. Two other 1 "weedy" genera, Rumes and 
Acalypha , were represented by one seed each, and four seeds were are possibly 
Euphorbia collata . These plants invade disturbed habitats, so their presence 
around plantation slave quarters was hardly remarkable. 

The Polygonum seed, 1.6 millimeters long, and trigonous with concave sides, 
was probab 1 y Pol ! y gonum hy dropi peroi des . This species inhabits swamp forests, 
streams, and ditches l Radford, Ahles, and Bell 1968), so its presence in an 
irrigation ditch (Feature F8 at 38BK245) is understandable. How it became 
carbonized is more problematic, but may be an indication of fires located 
outside of the domestic structures. Fires may have been used to clear areas 
of weedy growth, or the seed may have been dispersed into a fire used for 
some other outdoor activity such as boiling laundry, making soap, or burning 
rubbish. 

Several grass seeds were also found in the samples. Three carbonized seeds 
of goosegrass ( Eleusine indica L.) were of special interest. Goosegrass is a 
common grass in the Carol inas today (Radford, Ahles, and Bell 1968:116), but 
is a native of Asia (Martin 1972:19). The three seeds found in the early 
nineteenth century Yaughan Plantation samples (Appendix G) may be the ear- 
liest evidence of its occurrence in the New World. 

The seeds termed "unidentified type one" were the most numerous grass seed 
found and the most troubling. They are roughly cylindric with a beveled end 
and shallow groove along one side. Their classification as a grass seed is 
somewhat questionable, since the bevel and the shallow groove are on the same 
side of the seeds. The seeds are highly variable in size, ranging from 2.3 
millimeters to 4.2 millimeters long. It is possible that this seed type is 
not a grass, and may, in fact, not be a seed. 



292 



The other "unidentified grasses" category included one seed of either Setaria 
or Paspalum from the kiln (Structure 24510 at 38BK245. Identification could 
not be more definite, since the seed was both distorted and eroded. The 
other four seeds in this category were fragmentary remains of small grass 
seeds such as Panicum or Digitaria . Like the other weedy plants identified, 
the grasses were likely to have been colonizers of disrupted areas of the 
plantations, and their seeds were most likely carbonized accidentally rather 
than as a result of any intentional human utilization. 

A paleoethnobotanical study such as this one has much less chance of gaining 
information concerning cash crops. This is hardly surprising, however, as 
excavations centered on domestic areas are not likely to encounter evidence 
of the processing, storage or transporting of cash crops, since these acti- 
vities were probably conducted in areas of the plantation removed from the 
domestic structures. An expansion of the excavations to include other areas 
of the plantations might have detected archaeological evidence for particular 
cash crops, but more likely in the form of structures associated with their 
storage or processing than in remains of the plants themselves. It is in 
gaining information concerning the subsistence practices of the Yaughan and 
Curriboo slaves that this study has been most successful. Otto 
(1980:318-337) found that at Cannon's Point Plantation, St. Simons Island, 
Georgia, the slaves used proportionally less wild animal foods than planters, 
and thus one might expect wild plants to have played a less important role in 
the diet of the Yaughan and Curriboo slaves. This, in fact, appears to have 
been the case. Cultivated plants seem to have provided the overwhelming por- 
tion of the plant food eaten by the slaves, with wild plants providing only 
occasional dietary complements. Also, the range of plants utilized for food 
is quite small, with only seven, or possibly eight, plants being utilized; 
only maize and rice seemed to be of any great importance. 

The reasons for the highly focal adaptation of the plantation slaves can only 
be speculated upon. It is, of course, possible that the paucity of wild 
plants is more apparent than real. The slaves may well have exploited wild 
greens such as pokeweed ( Phytolacca americana ) or goosefoot ( Chenopodium 
spp.), or potatoes and other root plants which have not been preserved, and 
it is also possible that they grew cultivated greens such as turnip or mus- 
tard ( Brassica spp.). Wild plants may have been of little importance due to 
the adequacy of the cultigen derived diet, which gave no motivation to gather 
wild foods. On the other hand, the lack of wild plant utilization may have 
reflected the particular social conditions of the slaves. Effective exploit- 
ation of wild plants requires considerable mobility in order to visit the 
often dispersed locations where the plants occur, and considerable freedom to 
schedule activities so that one can gather the wild plants during their 
usually restricted harvest period. Since the documents have shown that the 
slaves probably did have some mobility, the lack of wild food sources may be 
a result of a conscious decision not to use wild foods, ignorance of which 
plants were edible, or restricted freedom to schedule activities. 



293 



The overall adequacy of the slaves' diet was difficult to assess. The pri- 
macy of rice and possibly corn as foodstuffs suggested a diet heavy in 
carbohydrates and low in other nutrients, but this conclusion must be tem- 
pered by the knowledge that other foods were probably eaten but not pre- 
served. It is speculated that the slaves' diet was constricted, not to the 
point of chronic malnutrition, but rather to that of culinary monotony in the 
vegetal diet. This arrangement would provide the plantation owners with a 
healthy and relatively inexpensive work force. 

Zooarchaeological Remains 

While the absence of a particular species may not be indicative of its ab- 
sence in the slave diet due to differential preservation, the undoubted 
presence of species does indicate a certain association with slaves and 
presumably with slave diet. Unfortunately, compilation of data on the zoo- 
archaeology of the sites was greatly hampered by poor preservation, which was 
of much greater extent than was the case with the ethnobotanical material. 
The following tables present data developed by Wood and Reitz (Appendix H) as 
well as by SSI staff members. 



TABLE 62. Animal Food Sources 

38BK75 38BK76 38BK245 

White tailed deer x 

Common oyster x x 

Quahog clam x 

Opossum x 

Goose x 

Freshwater catfish x x 

Cow xxx 

Pig x x x 

Dog x 

Snake/lizard x x 

x indicates presence 



294 



TABLE 63, 



Minimum Number of Individuals 
"3SBT75 



38BK76 



38BK245 



White tailed deer 

Common oyster* 

Qua hog clam* 

Opossum 

Goose 

Freshwater catfish 

Cow 

Pig 

Dog** 

Snake/lizard** 



Cow/pig 
Other 



1 





1 (present) 


3 


(present) 






1 


(present) 






1 


1 


1 








1 






1 


3 


2 




4 


5 


2 




6 




1 




1 




1 




1 


1 


6+ 




14+ 


8 


4 




10 


3 


2+ 




4+ 



*The oyster and clam shell were always found in close association with archi- 
tectural features and often in a mortar matrix. The actual minimum number of 
individuals would number in the hundreds if properly analyzed. It is felt, 
however, that little if any of the shellfish collected were included in the 
slave diet. 



**These are included since they could potentially have been food sources 
although it is more likely that they are not. 



Several striking facts are apparent in this data. Only pigs and cows are rep- 
resented at all three sites. While caution must be used in comparing rela- 
tive frequencies, the overall bone weight, fragment count, and MNI show pigs 
and cows outweighing and outnumbering the other resources combined, by a fac- 
tor of 2:1 or more. It was concluded that these domestic animals were a ma- 
jor, if not the major, meat source in slave diet at Yaughan and Curriboo. 

As with the botanical material, the absence of bone did not necessarily mean 
absence in the diet; however it was remarkable that no chicken, duck, or rab- 
bit bones were preserved, whereas goose was present (Table 61). Seven wild 
species and genus, including borderline cases (snake and dog), were represent- 
ed while only two domestic species were represented; however those were in 
high quantities of individuals. When a wide variety and low frequency of a 
certain class of resource are present, combined with high quantities in a 
second and restricted class of resource, it seems safe 
specialization in the former and a concentration in the 
of wild resources was not a specialized activity whereas 
cularly cow and pig, were more heavily used. 



to conclude a lack of 
latter. Exploitation 
the domestics, parti- 



295 



A further outstanding feature of the fauna! resources at the sites was the 
relative number of species present at the sites and the total bone weights 
when pH was taken into consideration. Although more features and units were 
excavated at 38BK76 than at the other sites, and although Site 38BK76 had the 
highest pH and, therefore, the best potential for preservation, only four 
species were identified there, as compared to five at 38BK75 and eight at 
38BK245. This disparity was also reflected in total bone weights. Site 
38BK76 had 248.16 grams, while 38BK75 had 320.80, and 38BK245 had 1364.01. 
If the original amounts of bone at the sites had been nearly equal and equal- 
ly distributed and the pH had been the same, 38BK76 would be expected to hav„ 
the greatest amount of bone and greatest variety since more excavation was 
conducted there. If all factors had been equal except the pH, then 38BK76 
could also have been expected to have had the most bone. It was concluded 
that there was either less bone at 38BK76 to begin with, or differential 
sampling resulted in a low bone count. The latter possibility can be dis- 
counted since virtually the entire site was excavated. This leaves the con- 
clusion that 38BK76 simply had less bone than the other sites, and its 
inhabitants consumed less meat in their diet. 

This could have reflected a difference in status and material comfort among 
the sites. Documentary evidence indicated that the owners of Curriboo were 
better off economically than those of Yaughan, and this appears to have been 
reflected in the slave population as well. This hypothesis is further cor- 
roborated by other differences noted in the architecture and artifact assem- 
blages discussed above and in Chapter XII. Although it is impossible to 
directly compare amounts of seeds and bones, the relative lack of bone at all 
three sites compared to other historic sites (Otto 1976; South 1977; and 
Garrow 1981) indicated that meat sources did not play a large role in the 
slave diet. As noted above, the vegetal diet was probably fairly monotonous, 
and without significant amounts of meat, the overall diet would have been 
just as monotonous. 

The apparent lack of meat in the slave diet ran counter to the idea that the 
gun parts at the sites reflected slaves hunting for their food. Three 
possibilities exist to explain the presence of the guns: that the overseer 
or owner used the guns exclusively, that slaves used guns exclusively, or 
that both owners and slaves used guns. The guns could have been used in one 
or more of a variety of situations. Documentary sources noted use of guns 
for hunting for food, killing or chasing pests attacking crops, and coercion 
on the part of the owners or overseers (Morgan 1977:42-43). To these may be 
added participation in armed conflict. The first two uses were often in the 
provenience of slaves, while the last two were not, in most cases. The most 
likely situation probably would have been slave or white use of guns for 
hunting or chasing away pests. Morgan (1977:42-43) points out that although 
the law prohibited more than one slave from using a gun for hunting per 
plantation in South Carolina, this law was often broken, and there was no 
limit on slaves having guns to chase away pests. It seems apparent that 
slaves did, indeed, use guns. The relative lack of wild faunal remains at 
the sites indicated that the guns were used to chase away pests and perhaps 
to provide wild game for the overseers or owners. 



296 



The idea that slaves had guns available to them and engaged in hunting for 
their masters is not new, as witnessed by the laws concerning gun by slaves 
use in the eighteenth century. However, the reemergence of this idea in the 
archaeological literature resulted in the hypothesis that wild game provided 
a mainstay in the slave diet. This hypothesis may have been influenced by 
Kenneth Stampp's (1956:284) remark that owners encouraged the slaves to 
"feast occasionally on wild game". Such an hypothesis runs counter to the 
majority of historical documents and the archaeological evidence provided by 
Otto (1977), Drucker and Anthony (1979), and now at Yaughan and Cum" boo. 

Conclusions 

In summary, although it is unfortunate that preservation and disturbance 
prevented meaningful comparisons with Otto's (1976) work at St. Simons and 
South' s (1977:179-182) bone ratio, some conclusions can be drawn concerning 
the subsistence of the slaves at Curriboo and Yaughan. 

1. The diet was primarily vegetal with the domestics, corn and rice, being 
the mainstays of a rather monotonous diet. 

2. The mainly vegetal diet was necessarily high in carbohydrates which could 
fuel hard manual labor, but lacked animal protein. 

3. The meat diet was of secondary importance and depended primarily on do- 
mesticated cows and pigs.- 

4. The overall diet was partially supplemented with minor amounts of wild 
food sources, both faunal and botanical. 

5. The variety and amounts of meat sources indicated a higher status or bet- 
ter material conditions for the slaves at Curriboo than at Yaughan. 

A note of caution should be added here. It has been noted that the data pre- 
sented on faunal remains, in particular, is potentially misleading as a re- 
sult of difficulties of preservation. We feel confident that with respect to 
these difficulties, we have not gone beyond the data. However, there is a 
potential difficulty which is not often stressed by historical archaeolo- 
gists. This difficulty is that of generalizing from individual sites. While 
it can be claimed, with some justification, that patterns can be detected in 
material culture, some parts of culture appear to lend themselves to general 
pattern studies better than others. A particular case in point is slave 
diet. 

The historical literature is rife with contradictory statements concerning 
what slaves ate, how they ate it and where, and who controlled their diet. 
This is amply illustrated by a perusal of Stampp (1956:282-289), Rose 
(1964:122 and 123), Handler and Lange (19797:86-89, 54, 73), Morgan (1977:42, 
47-50), Hilliard (1972 and 1969:5), and Miller (1978). It seems apparent 
that what slaves ate, how much, when, how it was prepared, and whether or not 
hunting was allowed or even encouraged depended upon individual masters. 
Some masters closely watched over slave diet providing (for the times) a 
balanced, but often uninteresting diet. Other owners did not. It is probably 



297 



safe to state that there were as many different slave diets as there were 
owners, at least in the eighteenth century. It could very well be that as 
the nineteenth century progressed, standardization on an elementary scale set 
in as a reaction to labor efficiency, agricultural journals, peer pressure, 
cotton growing, and even journalistic attacks by abolitionists. 

However, future researchers should carefully examine the results of this 
project, especially with respect to subsistence, before generalizing to all 
slave quarters. It may be that the pattern presented here will be a common 
one on other plantations, but the variability in food procurement and diet 
illustrated in the historical literature indicates this will probably not be 
the case. 



298 



XI. PLANTATION LIFE 

Introduction 

Since the publication of Kenneth Stampp's The Peculiar Institution (1956) and 
Stanley El kin's Slavery (1960), historians have wrestled with the problem of 
the nature and integrity of Afro-American culture. While the profession has 
by no means reached a consensus about antebellum black culture, major steps 
have been taken in the past decade with the publication of Peter Wood's Black 
Majority (1974), Stanley Fogel and Robert Engerman's Time on the Cross 
(1973), Eugene D. Genovese's Roll , Jordan, Roll (1974), and Herbert Gutman's 
The Black Family in Slavery and Freedom, 1 /bU— 1925 (1977). Despite serious 
disagreement among these scholars, aTT of them have endeavored to show the 
viability of black culture within the restrictions of the plantation system. 

This project makes a substantial contribution to this discussion because of 
its scope, dealing with eighteenth century, low-country plantations, and 
because of the nature of its evidence. The artifacts can be seen as direct 
black testimony, which differs substantially from conventional historical 
sources that are typically white descriptions of black behavior. In this 
context, the historical component has treated the following questions: 

1. What was the degree of cultural continuity in the slave quarters? 

2. How does the historical literature on slavery help interpret the docu- 
ments and the archaeology stemming specifically from the sites? 

3. How does the project as a whole extend current understanding of slavery? 

Question 1, which addresses cultural continuity, became the pivotal question 
for both the historical and archaeological components. The question of conti- 
nuity is critical in the historical perspective in order to ascertain from 
purely historical considerations and questions the degree to which there was 
an autonomous or distinctive slave community that responded to and interacted 
with the white family. At the same time, the point needed to be established 
since continuity was assumed in the archaeological investigation, which en- 
abled that inquiry to concentrate on relationships among the artifacts. Con- 
tinuity is a complex question, requiring elaboration of its meaning and iden- 
tification of factors that created, maintained or destroyed it. Partially in 
response to these abstract considerations, and partially in response to the 
documents that were available, the question was resolved into a series of 
subquestions. These included: 

1. What was the size of the slave quarters? 

2. What was the effect of purchase upon the slave community? 

3. What was the effect of inheritance practices upon the slave community? 

4. What was the effect of sale upon the slave community? 

5. What evidence was there of family and kinship bonds among the slaves? 

6. What evidence was there of perpetuation of Africanisms in the slave quar- 
ters from wri tten evidence? 

7. What evidence was there that slaves acted independently of their masters 
and were able to interact with whites on terms other than those that the 
whites dictated? 



299 



Plantation Life and Culture 

In a now classic definition of the Southern plantation system, Lewis Cecil 
Gray linked the evolution of the plantation to staple crop agriculture and 
forced labor (Gray 1933). In South Carolina, "forced labor" meant first 
Indian and then predominantly black labor. The demographic components of the 
history of St. Stephen's Parish were discussed in Chapter III. Data present- 
ed in that chapter showed that the proportion of Indians in the slave popula- 
tion decreased after 1708, when it peaked at one-fourth of the total slave 
population. The average number of slaves imported from the West Indies and 
from Africa continued to grow, reaching its peak in the period 1735-1739. 
After a slump in the 1740s, importation of Africans again accelerated in the 
1750s. The period during which the settlement of Yaughan and Curriboo Planta- 
tions took place thus coincided with the rapid expansion of Africans in the 
Carolina slave population. Although an occasional slave in the lists includ- 
ed in probate inventories was described as an Indian or more frequently as a 
Mustee , no information obtained thus far indicated that the constituency of 
these plantations differed radically from the mainstream, which was overwhelm- 
ingly African. 

The following section treats the history of the slave population of Yaughan 
and Curriboo plantations and deals with the issue of continuity. Eugene D. 
Genovese's insightful research has shown that black and white history are two 
sides of the dialogue that constitutes Southern history (Genovese 1974:2). 
Therefore, this section will begin by discussing the parameters of white cul- 
ture and will then proceed to a discussion of the black response and behavior 
within the limits imposed by the plantation. 

White Occupancy 

Evidence from the chains of title and the information on the family suggested 
that Yaughan was occupied in some way in the 1740s. Isaac Cordes* inventory 
listed tools and stock, but no slaves, at Yaughan in 1745 (Inventory of Isaac 
Cordes, 9 August 1745, Inventories, Vol. 67A, 1732-1746, pp. 316-332). When 
his son John, who inherited the plantation from his father, died in 1756, he 
left his entire estate including at least 51 slaves in the custody of his 
brothers-in-law (Will of John Cordes, Record of Wills, Vol. 7, 1752-1756, pp. 
582-584). The account book from the trusteeship of Samuel Cordes, who also 
owned Curriboo Plantation in these years, survives and refers at several 
places to Yaughan. 

Shortly after Samuel Cordes (d. 1796) took control of his brother-in-law's 
property and assumed guardianship of his children, he inventoried the estate. 
The inventory in 1764 listed 65 slaves, a growth of 14 slaves since the in- 
ventory in 1757 following John Cordes' death, and enumerated household goods 
at Yaughan worth sixteen pounds (colonial currency) (John Cordes Estate, Ac- 
count Book, 1756-1798, p. 6, CC). In December of that year, Samuel Cordes 
(d. 1796) credited indigo worth t 750 to the estate's account. Clearly, the 
plantation had been in production since John Cordes' death. It was an indigo 



300 



and possibly a rice plantation and had probably been in production in John 
Cordes' lifetime, i.e., prior to 1756. Evidently reviewing earlier accounts, 
Samuel Cordes (d. 1796) noted the purchase of "1/2 doz[en] broad Hoes for 
Youghan," amounting to four pounds, in May 1758 (p. 12) and the sale of indi- 
go worth b 984.14 "made at Yaughan" in December 1756 (p. 11). 

The plantation was in the daily supervision of an overseer and possibly had 
been since Isaac Cordes acquired it in 1742. In Isaac Cordes' inventory, the 
appraisers noted cows, calves, working oxen and horses at Yaughan and "Sun- 
dries in Company with Peter Lequeux" (Inventory of Isaac Cordes, 9 August 
1745, recorded 6 December 1749, Inventories, Vol. 67A, 1732-1746, pp. 
328-331). The Lequeux family settled in St. James, Santee, and later resided 
in St. Stephen (Will of Peter Lequeux, Record of Wills, Vol. 14, 1771-1774, 
pp. 107-110; Misenhelter 1977:6). The reap hooks, spades, axes and hoes 
listed in this portion of Isaac Cordes' inventory are separate from other 
listings and follow an enumeration of items owned jointly with Thomas Cordes 
(d. 1748) at Curriboo. The livestock at Yaughan was explicitly identified as 
belonging to the plantation; it was listed after stock listed at Curriboo in 
the same way that the listing of items held jointly with Peter Lequeux fol- 
lowed the list of goods owned jointly with Thomas Cordes (Inventory of Isaac 
Cordes, 9 August 1745, recorded 6 December 1746, Inventories, Vol. 67A, 
1732-1746, pp. 329-332). The structure of the document together with what 
was known about the individuals strongly indicated that the agricultural 
implements were located at Yaughan and that Peter Lequeux looked after the 
plantation as either a full -time overseer or from his own plantation in the 
neighborhood on a part-time basis. In 1762, Samuel Cordes paid Peter 
Lequeux, Jr., for nine months "overseeing at Youghan," which is additional 
evidence that the Lequeux family had a standing relationship with the Cordes 
family at their plantations in St. Stephen's Parish (John Cordes Estate, 
Account Book, 1756-1798, p. 14, CO. 

The twelve yoke of oxen listed as being at Yaughan in 1745 suggest that the 
plantation had not been developed to a great extent. When Thomas Cordes, 
Isaac Cordes' brother, died in 1748, he left his sons six oxen each, together 
with plantation tools and a plantation. This implies that six oxen were es- 
sential to running a plantation. Having twice as many oxen, which were used 
as draft animals in clearing and ploughing the land, on Yaughan indicates 
that this was a "frontier" plantation. The absence of slaves also adds to 
the image of a "frontier" plantation that had been recently acquired and was 
in the early phases of operation. Assuming that the tools held with Peter 
Lequeux were being used on Yaughan confirms this impression but indicates 
that there were slaves on the plantation at least part of the year. Sup- 
porting the impression that this was a newly settled plantation is the fact 
that there were 2 new spades, 3 new axes, 10 old axes, 6 new narrow hoes, 6 
new broad hoes, 1 new fanner and 1 new rice sieve listed in the inventory. 
There were in addition 22 reap hooks. The number of new implements suggests 
that the plantation was recently settled as does the number of spades and 
axes, which were used to fell trees. There were, in fact, more axes than 
hoes, which were a critical implement in eighteenth century agricultural 
practices. The configuration of implements, moreover, implies clearly that 
this was more than a forested tract on which cattle were run. The numbers of 



301 



tools indicate that there were a minimum of between 12 and 22 slaves of 
working age resident on the plantation at least part of the year. The rice 
sieve and hooks suggest plans for rice cultivation, either future (within one 
growing season), or for a crop already in the field. 

After 1760, there is regular evidence describing events at Yaughan. In 1761, 
Samuel Cordes recorded payment for "making an Oven at Youhan" (John Cordes 
Estate, Account Book, 1756-1798, p. 14, CC); in 1769, he paid for "delivering 
four Wenches in Child bed at Yaughan" (p. 46), and in 1770, he had a "Chimney 
at Yaughan" rebuilt (p. 50). Between 1762 and 1774, he paid overseer's wages 
to five men: Peter Lequeux, Jr., Jonathan Dubose, Enoch Linerieux, Isaac 
Couturier and Isaac Barnes (pp. 14, 16, 24, 26, 46, 52, 55, 59). Four of 
these were members of old Huguenot families who had settled in St. Stephen 
and the fifth, Isaac Barnes, died resident in the parish. Samuel Cordes was 
one of the executors of Barnes' estate when he died in 1784. Barnes owned 
property worth over 400 pounds. This challenges the conventional image of 
the overseer as a servant of the plantation owner, wholly without property of 
his own. In January 1772, Isaac Barnes was paid "for one share of his in 
Negroes that Crop" implying that like Peter Lequeux in 1745, the overseer at 
Yaughan had made an investment in the running of the plantation (p. 55). 

On May 5, 1774, Thomas Cordes acknowledged receiving his share of his inheri- 
tance from the custody of his uncle, Samuel Cordes (John Cordes Estate, Ac- 
count Book, 1756-1798, p. 73, CC). There are no further references to over- 
seers at Yaughan, and since Thomas Cordes formally purchased Yaughan from his 
elder brother John on May 10, 1775 (John Cordes to Thomas Cordes, Release of 
3 Tracts of Land, 10 May 1775, recorded 7 April 1786, Deed Book R-5, pp. 
193-196, RMC), this indicates that Thomas Cordes occupied the plantation and 
worked it himself. He stood godfather to Peter Porcher (b. April 10, 1777) 
in St. Stephen's in 1777 (John Cordes Estate, Account Book, 1756-1798, p. 
190). In April 1778, the vestry of St. Stephen elected him churchwarden, a 
lesser parish office, but one, nonetheless, given to residents of the parish, 
and in 1785 he was elected to the vestry itself (Misenhelter 1977:56-57). 
Thomas Cordes married Charlotte Evance in 1784; he was then a member of the 
(state) House of Representatives, which was also an office conventionally 
reserved for influential men (Richardson 1942:152), and which also required 
residence in the parish. He was elected to the state constitutional conven- 
tion in 1790, all of which adds to the impression of Thomas Cordes as a sub- 
stantial resident of St. Stephen's Parish. 

It is clear that Thomas Cordes participated actively in the Revolutionary 
War. Evidently, however, he did not leave St. Stephen's entirely. After the 
war, he apparently returned to the parish, where he settled with his wife and 
began to raise a family, again beginning the elite life of a planter. In 
1785, he purchased a barrel of rice from Hezekiah Maham, a neighbor (H. 
Maham, Ledger, 1765-1794, p. 44, USC), and in 1787, he borrowed 1 = bushels of 
indigo seed from John Fitzgerald, another neighbor (Palmer Ledger, 1777-1807, 
p. 7, USC). Like other planters in St. Stephen, he seems to have tried to 
rebuild the old eighteenth century plantation on the basis of rice and 
indigo. It was not a wise decision and ultimately led to the sale of slaves. 
Thomas Cordes recognized that he was in bad straits, since in 1800, he sold 
15 slaves to his sister-in-law Margaret Cantey with the stipulation that she 
hold the slaves in trust for his wife (her sister) Charlotte. When he died, 



302 



he still owned 47 slaves (Inventory of Thomas Cordes, 22 June 1807, Inven- 
tories, Book D, 1789-1810, p. 429, SCDAH). He had inherited 20 slaves from 
his father's estate (John Cordes Estate, Account Book, 1756-1798, p. 32, CO, 
so he clearly enjoyed a modest prosperity. Assuming that all of the 47 he 
owned when he died were alive in 1790 when he began to sell slaves, he held 
at least 86 slaves during his lifetime. It was not an inordinately large 
plantation, but according to Philip Morgan's estimates of the sizes of plan- 
tations in the low country, neither was it a particularly small one (see be- 
low, "The Black Community," for a discussion of sizes of plantations). 

A document parallel to the Cordes account book detailing Yaughan is not avail- 
able for Curriboo. There is other information in the accounts describing 
agricultural purchases and expenditures that is not linked specifically to 
Yaughan. Since this is data on the totality of John Cordes' estate, which 
included more property than Yaughan in both St. Stephen's and St. John's, 
Berkeley, this information can be construed to describe mid-eighteenth cen- 
tury plantation life in the Santee-Cooper region. It thus applies equally to 
Curriboo as to Yaughan. 

Samuel Cordes was one of the founders of St. Stephen's Parish in 1754 and he 
was commissioned to make brick for building the parish church. The vestry 
rejected the bricks, considering them of inferior quality (Misenhelter 
1977:6, 8). Samuel Cordes clearly resided in St. Stephen's on Curriboo, 
which he had inherited from his father in the 1750s, and he had possibly 
moved to the plantation as early as 1748, when his father died. The decision 
of his nephews John and Thomas Cordes to occupy plantations in St. Stephen's 
in the 1770s reflects the importance of kinship in determining patterns of 
residence and reinforces the impression that Samuel Cordes was continuously 
in residence in St. Stephen's Parish from the 1750s at least until the Revo- 
lution. In 1774 and 1789, however, he inherited significant holdings in St. 
John's, Berkeley, and when he died in 1796, he identified himself as residing 
in St. John's (Will of James Paul Cordes, Record of Wills, Vol. 18, 
1776-1784, pp. 203-205; Will of James Cordes, Record of Wills, Vol. 23, Book 
B, 1786-1793, p. 414, SCDAH). Thus, in the mid-1 770s, he may have begun to 
divide his time among his several plantations. 

John Cordes' estate produced indigo, rice, pitch, corn, beef and peas. This 
pattern of mixed agriculture is consistent with what is known about other 
plantations in this area. In the 1770s, for example, Henry Ravenel of Han- 
over, in St. John's, Berkeley, produced beef, corn, indigo, rice and naval 
stores (Henry Ravenel, Ledger, 1760-1774, SCHS). During the Revolutionary 
War, members of the Cordes family supplied beef, corn and livestock to Fran- 
cis Marion's and Hezekiah Maham's troops (Chapter V), and in the 1780s, Maham 
and members of the Palmer family continued to grow rice and indigo as well as 
corn, peas and oats (H. Maham, Ledger, 1765-1794; Palmer Ledger, 1777-1807, 
USC). The inference is plain, then, that Yaughan and Curriboo in the second 
half of the eighteenth century were primarily rice and indigo plantations 
with secondary investments in such subsistence crops and products as beef, 
hides, naval stores, peas and corn. Yaughan was smaller than Curriboo, hous- 
ing perhaps 80-90 slaves at its peak between 1785-1790. Curriboo was more 
prosperous; when Samuel Cordes died, he had 103 slaves at Curriboo, almost 
one-fourth of his total of 408 slaves (John Cordes Estate, Account Book, 
1756-1798, pp. 138-141). 



303 



At his father's death, Thomas Cordes, Jr. lived at Mil ford Plantation in St. 
Stephen's where he worked 82 slaves (John Cordes Estate, Account Book, 
1756-1774, pp. 141-146, CC). Samuel Cordes explicitly willed Curriboo to his 
son Thomas, Jr., who evidently left Mil ford to live at Curriboo. He survived 
his father only three years, and when he died, he left his house in Pineville 
to his widow Rebecca (Will of Thomas Cordes, Jr., Record of Wills, Vol. 27, 
Book C, 1793-1800, pp. 504-507). He had apparently already begun to live 
part of the year in the village, and his widow evidently left the plantation 
entirely. She owned the Pineville property and was living in Charleston at 
the time of her death (Will of Rebecca Cordes, Record of Wills, Vol. 43, 
1839-1845, pp. 680-683, CCPO). Her son James Jamieson Cordes left South Caro- 
lina permanently in 1821, and probably her son-in-law John Harleston, who had 
married her daughter Elizabeth in 1819, looked after the family property, 
since he owned plantations in St. Stephen's. 

Charlotte Cordes also went to live at least part of the year in Pineville 
after Thomas Cordes' death in 1806. In 1814, she sold 28 slaves, and al- 
though she and the four of her children who lived in St. Stephen's in 1825 
owned a total of 77 slaves in that year, the household at Yaughan itself may 
have already begun to break up by the mid-1 820s (Comptroller General, Tax 
Returns, 1824, St. Stephen's Parish, SCDAH). 

Economics of the Plantation 

Philip Morgan argues that the decade of the 1740s constituted a watershed in 
the development of the plantation in St. John's, Berkeley. After the hard 
times of the decade, he argues, the planters consciously tried to promote 
self-sufficiency within the plantation (Morgan 1977:27-31). Morgan has, 
however, minimized the extent to which planters supplied each other in a 
parish-wide network of self-sufficiency. A network of local exchanges 
emerged within which the planters supplied each other with essential goods, 
ranging from items for the household to extra bushels of corn for the stock 
and seed. Clearly Thomas Cordes participated in such a circuit, since he 
obtained rice and indigo seed in the 1780s from his neighbors. These 
accounts also document purchases by Samuel Cordes, John Cordes and Francis 
Cordes (H. Maham, Ledger, 1765-1790, p. 44; Palmer Ledger, 1777-1809, pp. 6, 
7, 50, and 74, USC), suggesting that the restricted scope of the exchanges 
meant that these transactions took place within the familiar matrix of neigh- 
bors and kindred. 

Systematic analysis of Henry Ravenel's Ledger between 1760 and 1771 details 
this local commerce (Henry Ravenel , Ledger, 1760-1774, SCHS). In these 
years, he traded with six neighbors: William Moultrie, Daniel Ravenel, James 
Ravenel, Samuel Richebourg, Samuel Williams, and Stephen Mazyck. Two, Daniel 
and James Ravenel were kindred, and two, William Moultrie and Stephen Mazyck 
were in-laws. He dealt in tar, pitch, hides, beef, small quantities of indi- 
go, tallow, bark and the rental of slaves. None of these items were part of 
the rice-indigo basis of the colonial economy, but all related to the daily 
functioning of the plantations. Kinship clearly informed these exchanges, 
and the picture that emerges reinforces the Insular quality of life in the 
rural parishes. 



304 



The John Cordes Estate Account Book presents a better picture of the econo- 
mics of the plantation in its relationship with other plantations and in its 
relationship with the larger economy. The book lists 112 transactions be- 
tween 1756 and 1774. Of these, 68 list the name of the firm or the indivi- 
dual. Thirty-nine of these transactions took place between the estate and 
individuals and the remainder were between the estate and a firm in Charles- 
ton. In these exchanges with Charleston firms, six of the seven firms whose 
identity is known were also involved in the overseas slave trade (Higgins 
1964:205-217). Exchanges with these firms concerned almost exclusively rice, 
indigo and corn in large quantities, and rarely did they involve small consum- 
ables (e.g., several bushels of corn or bottles of wine and rum) for the plan- 
tation. These firms acted as bankers for the members of the family. In 
1767, Samuel Cordes noted the proceeds of t 350 paid to the estate's account 
with Livingston & Champney for the sale of one crop of indigo. Over 1200 
pounds had been carried over to the estate's credit with the firm from the 
preceding year (John Cordes Estate, Account Book, 1756-1798, p. 26, CO. 

These commercial transactions involving the plantation's viability over the 
long run were conducted in a separate circuit from those involving daily pro- 
visioning. As the analysis of the Ravenel Ledger implied, these took place 
almost exclusively among neighbors and kindred. There were 39 transactions 
with 25 individuals. They ranged from one-time exchanges to multiple ex- 
changes, the most frequent of which was four (Table 64). These relation- 
ships, then, were fluid so that no permanent relationship existed between one 
or two individuals in the parish. Nine of these 25 men were neighbors, mean- 
ing that they were residents of St. Stephen's Parish; one was a kinsman, and 
two were in-laws (Table 65). Nearly half (12/25) were exchanges between men 
who already knew each other and already shared common bonds. As had been the 
case with kinship, marriage and locale, these economic exchanges served to 
reinforce the restricted scope of contacts and to heighten the insularity of 
the parish. 

In 29 transactions, both the supplier and the item exchanged were known. In 
this analysis the two circuits become extremely clear (Table 66). Items 
having to do with consumption and provisioning the plantation, including sup- 
plying the slaves with goods, were almost entirely local. Eight of the nine 
exchanges of Negro goods involved neighbors, kindred or in-laws. The two ex- 
changes involving poultry were between neighbors, and six of the nine transac- 
tions which involved primarily small quantities of beef or an odd steer were 
between the estate and neighbors. By contrast, the commercial exchanges, 
which were concerned with exports from the colony, did not involve neighbors. 
One exchange did take place between the estate and Samuel Prioleau, a mer- 
chant in Charles Town who married one of John Cordes' daughters. The pattern 
of marriages in the family leads to the conclusion that this was more or less 
an arranged marriage between Catharine and a merchant with economic links to 
the family. 

Slaves occupied an ambiguous position in these circuits. On the one hand, 
six of the seven firms with which the estate dealt were also involved in the 
slave trade, implying clearly that the Cordes were in a position to acquire 
slaves through the Charleston markets. Since the acquisition of labor was 
critical to the survival of the plantation, obtaining slaves in this sense 



305 



TABLE 64. JOHN CORDES ESTATE, ACCOUNT BOOK, 
FREQUENCY OF TRANSACTIONS WITH INDIVIDUALS, 1756-1774 



Number of Exchanges Per Individual 



Frequency 


% 


Cumulative % 


17 


68 


68 


3 


12 


80 


3 


12 


92 


2 


8 


100 



25 



100.0 



100.0 



Source: John Cordes Estate, Account Book, 1756-1798, CC. 



TABLE 65. INCIDENCE OF BONDS AMONG PARTICIPANTS IN THE EXCHANGES, 
JOHN CORDES ESTATE, ACCOUNT BOOK, 1756-1774 





Frequency 


% 


Cumulative % 


Neighbor 


9 


36 


36 


Kindred 


1 


4 


40 


In-Law 


2 


8 


48 


Unknown 


13 


52 


100 



25 
Source: John Cordes Estate, Account Book, 1756-1798, CC. 



TABLE 66. CROSS TABULATION, SELECTED ITEMS* BY 
INDIVIDUAL/FIRM, JOHN CORDES ESTATE, ACCOUNT BOOK, 1756-1774 



Item 



Total 



Nei ghbor/Ki n/I n-Law 



% 



Other 



Negro Goods 


9 


Beef 4 Steers 


9 


Poul try 


2 


Rice 


1 


Indigo 


4 


Corn 


2 



8 
6 
2 






88.9 

66.7 

100.0 



1 

3 


11.1 
33.3 



1 
4 
2 


100.0 
100.0 
100.0 



* 2 items, 1 pair of shoes for a member of the family and 1 payment of 
commission were deleted. 



Source: John Cordes Estate, Account Book, 1756-1798, CC. 



306 



was analogous to marketing indigo, rice and corn. On the other hand, provi- 
sioning slaves with shoes took place within the local circuit of exchange, 
implying that slaves, once acquired, were treated as inhabitants of the plan- 
tation that in many ways resembled a small village, whose care was a matter 
to be resolved locally. By implication, then, once on a plantation, a slave 
became nested in an insular mentality and his or her contacts were restricted 
to a relatively small area. 

An analysis of Henry Ravenel's Day Book offers a closer look at the internal 
functioning of the plantation (Henry Ravenel , Day Book, 1763-1766, SCHS). 
His ledger and the John Cordes Estate Account Book reflect the ebb and flow 
of credit for the plantation as a whole. Figure 87 represents three-month 
moving averages for the frequency of expenditures by month, aggregated for 
the period 1756-1774. Expenditures peaked in frequency of record in Decem- 
ber, suggesting that at the end of the year, the plantation master sat down 
with his numerous bills and receipts and brought his account to date, cal- 
culating his expenditures against the credit he had obtained as a result of 
the sale of the crops, which had been brought in by the end of November. 
This pattern does not necessarily reflect the time when the slave actually 
received the items or when the plantation itself may have received or sold 
the items. 

Ravenel's Day Book, on the other hand, was a private record that he kept on a 
more frequent basis detailing mainly small transactions, principally within 
the Hanover Plantation itself. There is some overlap with the Ledger, but 
the exchanges noted in the Day Book largely concern Ravenel's trading with 
his slaves. These transactions were exclusively in cash and were very small; 
the mean value was k 1, 6s. Transactions seem to have involved slaves on 
neighboring plantations as well as those from Hanover. The slaves purchased 
relative luxuries, including flannel and rice, and appear to have supplied 
the product of their own husbandry, skill or ability to forage (Table 67). 
Sixty-three percent involved sale of fowl, hogs and corn; 18.5 percent in- 
volved supplying skills (i.e., mending a chair or table) or a product of 
skill (making a basket or a tub). Finally, 18.5 percent reflected the 
slaves' ability to exploit the environment for fish, honey or wood. 

Other records of plantations in St. Stephens and St. John's, Berkeley echo 
this pattern. The whites supplied certain basic items, such as blankets, 
shoes and corn, but this was not exclusively a provisioning, and the slaves 
elaborated on these staples. Thus, one of the Palmers noted "A list of Ne- 
groes took out Blankets in December 1796," the verb "took out" implying an 
element of slave participation rather than a strict distribution of supplies 
by the whites (Palmer Ledger, 1777-1809, p. 1, USC). Thomas Walter Peyre, of 
St. Stephen's, on the other hand, recorded giving out corn for the horses on 
May 17, 1843, for Negroes on May 29, 1843, and for hogs, fowls, and horses on 
June 5, 1843 (Thomas Water Peyre, Journal, 1834-1850, p. 205, SCHS) on the 
same page of his ledger reserved for "Corn Used from the Corn House." Peyre 
may have been particularly systematic, but an unidentified member of the 
Ravenel family commented in April 1830 that he planted cotton at Pooshee and 
"gave out 560 yards of colored homespun to the negros for summer clothes" 
(Thomas Porcher Ravenel, Papers, Crop Book for Planting, 1830-1832, SCHS). 
Slaves, however, evidently made their own clothes from the dry goods sup- 
plied, and also augmented their diet. In 1833, Ravenel carefully recorded 



307 



u 

o 

c 
o 



c 
o 



o 



4) 

> 











2 








c/i 












^ 








i/ 












*^s 


\ 

V 






1 \ 












*^^s 


f 






\ i 












* 








M 














N 






I 
\ 
t 














• 

• 






/ 












• 








/l 

/ 












\ 








1 

\l 












1 


1 






/ 












/' 








f 
1/ 










,'' 




















7 

| 










/I 

\ 

1 1 








/ / 
/ f 












\i 








■vO* 


S*X 










\ J 

A 
/ 1 












xS. 
\X 
x^ 

X 


vv 






l \ 














\ 
\ 








I s 

























o 

i/) o 

c ca 
o 

— >> 

•PC A 

u rs. O 

TO Ps 

(/> t— i i/) 

C I - 

ro U3 i — 

s.m ai 
i — I*- c 

^ a; 

ps 4- > 
co o -C TO 

•M OS 

Lui - C 

Q£ OJ o >> 

=> -Q £ s- 
C3 S C 

m 3 >i«J 

ll. z: co x 



e 
u. o 

E 



> 
o 
3 



3 O 

> i 



CO o 



<s> 



o 



to 



O 
fO 






o 



m 



m 



°s 

>» o 

e u 

• o 
a •» 

O" c 

• o 



308 







TABLE 67. 


Goods and Serv 


ices 


Supplied 








Henry Ravenel 


of Hanover by 


Slaves, 1763-1766 










Cumulative 






Cumulative 


Item 




Frequency 


Frequency 




Percent 


Percent 


Fowl 




10 


10 




37.0 


37.0 


Hogs 




4 


14 




14.8 


51.8 


Corn 




3 


17 




11.1 


62.9 


Mending 


Chair 




18 




3.7 


66.6 


Mending 


Table 




19 




3.7 


70.3 


Bricklayer 




20 




3.7 


74.0 


Tub 






21 




3.7 


77.7 


Basket 






22 




3.7 


81.4 


Catfish 






23 




3.7 


85.1 


Honey 






24 




3.7 


88.8 


Trees 






25 




3.7 


92.5 


Rails 




2 


27 




7.4 


99.9* 


Total 




2/ 


27 




100.0 


100.0 



* Error due to rounding. 

Source: Henry Ravenel, Day Book, 1763-1766, SCHS. 



309 



the meat that slaves kept in the plantation smokehouse. He listed both debts 
he owed them for fowls he had taken and debts they owed him for meat they had 
"purchased." In still another fragmentary list, he recorded "Money due me 
from Negroes for items purchased," which included cards, a waistcoat, ker- 
chiefs, calico and the entry, "Lucy at Ophir for a lock" (Thomas Porcher 
Ravenel , Papers, "Money due me from Negroes for articles purchased," June 
1829-1833, SCHS). 

The late eighteenth century pattern manifested in Henry Ravenel 's Day Book is 
amply reflected in these additional documents. The uneven dialogue between 
master and slave illustrates the cruel contradiction that David Brion Davis 
has argued lay at the center of slavery, namely, that the slave was simul- 
taneously chattel property and human being (Davis 1964:60-62). The whites 
proved unable to ignore the human in the interests of the property, and tacit 
recognition of black humanity crept into exchanges on the plantation. The 
substance of these exchanges, the Ravenel papers show, suggest that part of 
the plantation's self-sufficiency was the result of the slaves' input , apart 
from their forced labor in the fields. Secondly, the demands of the plan- 
tation exercised the slaves' abilities and elicited responses that might have 
perpetuated Africanisms (i.e., basket-making) and fostered the slaves' sense 
of community within and identification with the plantation. 

The Impact of Inheritance and Sale 

Occupancy and provisioning were not the only white behaviors that circum- 
scribed black experience. Inheritance strategies, linked obviously to oc- 
cupancy, and sale also affected the formation of a slave community, and with 
the community the conditions which contributed to maintaining African prac- 
tices. Wills and inventories have survived for four male members of the 
Cordes family: Isaac Cordes (d. 1745), his son John Cordes (d. 1756), 
Isaac's brother Colonel Thomas Cordes (d. 1748), and his son Thomas (d. 
1763). Additionally, Henrietta Catherine Cordes' (d. 1765) will and inven- 
tory are available. These documents are extremely useful because they form a 
twenty-year unit of related individuals. The John Cordes Estate Account Book 
also supplies relevant information on inheritance practices and their impact 
on slaves as well as useful data on sales that took place within the family. 
Finally, bills of sale in addition to Samuel Cordes' (d. 1796) will and inven- 
tory, Thomas Cordes' (d. 1799) will and Thomas Cordes' (d. 1806) inventory 
have survived, constituting a block of documents for the turn of the century. 
A list of slaves was also attached to Elizabeth Cordes' (Samuel Cordes' grand- 
daughter) marriage settlement of 1819. 

One of the great advantages of having both wills and inventories for the same 
people is that the terms of the division of property are known as well as a 
precise account of the testator's personal property. This made it possible 
to determine population at risk and to provide a check on the record linkage 
of slaves from inventory to inventory c All of this information has been 
tabulated in Table 68, and since the same procedure was followed on each pair 
of documents (will and inventory), only the sequence followed in analyzing 
Isaac's relationship with John will be detailed. 



310 



a. 
o 



1 




41 




<-> 








u 




41 




£ 


a 


e 


4-> 




o 




>— 


o 




z 






4) 


•a 


a> 


01 




*-> 


o -o 


•^ 


H 41 


u 


•w 


41 


-* •*» 


A 


Vt '_ 


e 


«•- 41 


»* 


ee ^ 




c 




■*-> — 


o 


■< 


z 






o* 




z 


•o 


41 


*» 




1_ 


41 


£ 


C 


""* 


a 


Z 




-* -o 




</» c 


t> 


— 3 


e 


ae a 


3 


u. 


o 


4-> 


iu. 


«C 41 




GO 


a 


d o 


z 


ze-t— 


in 


• 41 "O 


o > e 


Z <9 3 


— o 


(/i ik 


.* 


</i i/i 


41 — (. 


> QC b. — 


a 4i 4> 


p— ** a. i 


io ^ 


jt 


Wl <A T3 


41 — 4> 41 


> a: aa — 


■a i— 


— ■*-» o — 


1/1 < 1— S 


in 


— 41 


* > 


■w <o 


o — 


h- </l 


41 


** 


m 


a 


i 


a 




z 



3 



Wl 

U 4) 

« "O 

i/> o 
— u 



f 



41 </» </» 

e « 4i 

o e "« 

i— o b. 

© -C D 

O I— O 









o o 

"3 O 






^K 



= "a 
o v. 

s. o 



g 



*^ 



41 



. </l 



- -mi 
v. .e-o 

C ■" b. 
41 « O 



"3 O 

3 •- 



4) X 

S <a 41 

" — J 

I 

•CO 

vO — *> 

in i 

!-■» U _* 

— U >/> 

£ — 



41 >/» 

— — <si 

> 
i/> E « 

4> O •— 

■a l «i 



e > «« 

.c 'o o 
o .- •— 



O 3 S 



< — < 



311 



TABLE 68. (Continued) 



Sources: Will of Isaac Cordes, recorded 9 August 1745, Record of Wills, 
Vol. 5, 1740-1747, pp. 406-409, SCDAH. 

Inventory of Isaac Cordes, 9 August 1745, Inventories, Vol. 67A, 
1732-1746, pp. 316-332, SCDAH. 

Will of Thomas Cordes, 25 April 1748, recorded 21 April 1749, 
Record of Wills, Vol. 6, 1747-1752, pp. 141-145, SCDAH. 

Inventory of Colonel Thomas Cordes, 21 April 1749, Inventories, 
Vol. B, 1748-1751, pp. 124-129, SCDAH. 

Will of John Cordes, 20 June 1756, recorded 3 December 1756, Vol. 
7, 1752-1756, pp. 582-584, SCDAH. 

Inventory of John Cord[e]s, 22 January 1757, Inventories Vol. S, 
1756-1758, pp. 22-28, SCDAH. 

Will of Thomas Cordes, 22 May 1762, recorded 6 July 1763, Vol. 
10B, 1760-1767, pp. 450-452, SCDAH. 

Inventory of Thomas Cordes, Undated, Inventories, Vol. V, pp. 
492-494, SCDAH. 

Will of Henrietta Catharine Cordes, 15 May 1760, recorded 31 
December 1764, Record of Wills, 1760-1764, pp. 442-444, SCDAH. 

Inventory of Henrietta Catharine Cordes, 10 January 1765, Vol. W, 
pp. 221-222, SCDAH. 



312 



Isaac Cordes left a slave to each of his three daughters, and the remainder 
of his personal property he divided equally among his son and three 
daughters. The appraisers listed 114 names and prices in the inventory of 
Isaac Cordes' estate and an additional 11 slaves held with the Curriboo pro- 
perty. Isaac Cordes owned 117 slaves outright and perhaps as many as 128. 
Since there is no mention of the Curriboo property in the will, and since the 
slaves named in the will did not appear in the inventory, the population at 
risk to be bequeathed, divided among the four children, was 114. The popula- 
tion at risk to be inherited by any single heir was 114/4 or 28-29. 

John Cordes (d. 1756) died 12 years after his father. His estate, when ap- 
praised in 1757, included at least 51 slaves, although some ambiguities 
suggest that the number was slightly higher. Three were clearly children 
(Little Janny, Little Grace and Little George). Linking names from Isaac's 
inventory to John's and checking price as an indicator matched 25 names. 
This empirical exercise came extremely close to the predicted estimate of 
28-29, which did not take mortality or relative values into account; the 
appraisers divided slaves so that values were equal, not the size of the lots 
of slaves. John Cordes also inherited slaves from his father-in-law Thomas 
Cordes by right of his wife (who was also his first cousin). Six slaves were 
a possible match. Since three slaves were apparently children, 48 slaves 
were at risk to have been i nheri ted in the population of John Cordes' estate 
in 1757". TTnrEy-one slaves were possible links with other records, and there- 
fore 31 out of 48 slaves at risk were inherited slaves, or 64.6 percent. 

This procedure was replicated for each documented relationship: from Colonel 
Thomas Cordes to his son Thomas, from Colonel Thomas Cordes to his wife Hen- 
rietta Catharine; and from Isaac Cordes to his brother Colonel Thomas with 
regard to the slaves held at Curriboo. Seven of the 11 slaves known to have 
been at Curriboo in 1745 show up in Colonel Thomas Cordes' estate in 1748. 
Seven of the eight slaves identified in Colonel Thomas Cordes' will appear in 
the inventory of 126 slaves. The total number in the estate was 127, but the 
population at risk to be bequeathed according to the terms of the will was 
only 119 because eight were earmarked specifically to individuals. He left 
1/6 of his personal property to his wife and the remainder in fifths to his 
five children. Between 19 and 20 slaves were at risk to appear in his wife's 
inventory, and 20 were at risk to show up in his son Thomas' inventory. 

In 1763, Colonel Thomas Cordes' son Thomas died, leaving 76 slaves. Of 
these, seven were clearly children, and therefore, the population at risk to 
have been inherited was 69. Twenty-one slaves were a possible match, which 
exceeds the predicted size, although it is close. There are several possible 
explanations. Thomas may have inherited a larger number of children from his 
father, and, although the value of his portion would have equalled the values 
of his siblings' lots, there would have been a larger population at risk. 
Slaves tended to name their children for kin, frequently fathers (Gutman 
1976:190-201), and therefore, the linkage may have matched too many names, 
assuming then that the linked name was that of a child and the name to which 
it was linked in the 1745 inventory belonged to a slave who went to another 
lot. At any rate, clearly the empirical linkage and the prediction are suf- 
ficiently close to validate the method as a means of assessing relationships 
between lists of slaves, although the statistics must not be construed as a 
precise measurement of actual fact. Slightly more than 30 percent (30.4%) of 
Thomas Cordes' (d. 1763) slaves were inherited. 



313 



Henrietta Catharine Gendron Cordes' estate was less complicated. She owned 
30 slaves when she died in 1765, and was in a position to have inherited 
property from her father as well as from her husband, Colonel Thomas Cordes. 
Considering only the slaves that she may have inherited from her husband, she 
stood to inherit a possible 19-20 slaves. Eighteen names match, so that 60 
percent of her slaves were inherited. 

These relationships can be viewed in two directions. On the one hand, the 
continuity from the testator to the heir can be measured. Out of all of 
Isaac Cordes' slaves, how many were tracked to later owners within the 
family, and by implication, how many disappear, presumably sold? On the 
other hand, the numbers can be construed from the perspective of the heir. 
Of his/her total estate, how many were inherited, and how many did he/she ac- 
quire from potentially distant sources? The earlier analysis indicated that 
members of the Cordes family dealt with firms in the overseas slave trade and 
hence the conduit with Africa is undeniably linked with their plantations. 

The concept of population at risk again comes into the analysis. Since not 
all heirs were considered, the total population at risk to be divided is not 
equal to the population at risk to have been inherited, when considering the 
situation from the testator's perspective. Thus, a total of 32 slaves be- 
longing to Isaac Cordes, including those he had an interest in at Curriboo, 
were discovered in either Colonel Thomas Cordes' inventory (7) or in his son 
John's inventory (25). Using the assumptions outlined above, that 114/4 was 
the population at risk per heir enumerated in Isaac Cordes' will, and that 
the slaves held at Curriboo were handled separately between the estate and 
Colonel Thomas Cordes, the population at risk to reappear in the documents at 
hand (Colonel Thomas Cordes' inventory and John Cordes 1 inventory) was 
29 + 11 =40. Of these, 32, or 80.4 percent, resurfaced in later documents; 
this is an extremely high degree of continuity from the perspective of the 
testator's slaves. A similar procedure involving Colonel Thomas Cordes' 
estate shows 92.8 percent of the 42 slaves at risk to appear in Henrietta 
Catharine's and Thomas Cordes' Inventories do, in fact, match. 

Viewing the transactions in the context of the three heirs' estates presents 
a somewhat different picture. In John Cordes' total estate, slaves known to 
have been inherited constituted 64.4 percent of his slave population, at risk 
to have been inherited. In Thomas Cordes' estate, they constituted 30.4 per- 
cent of his slaves, and in Henrietta Catharine Cordes* estate, they consti- 
tute 60.0 percent of her slaves. A series of factors may have affected the 
relationship; these were time elapsed between inheritance and death of the 
heir, relationship between testator and heir, and size of the heir's estate. 
Logically, one would expect the heir who survived the testator by the 
greatest number of years (Henrietta Catharine, surviving her husband, Colonel 
Thomas Cordes, by 17 years) to have the lowest percentage of inherited 
slaves; in fact, she had the highest percentage. As a woman and a widow, she 
was, however, in the least advantageous position to pursue aggressively a 
planter's life, and she also had the smallest number of slaves in her estate. 
"Relationship," a nominal designation which can not be analyzed in an 
equation with interval -scale data without a computer programmed with Multiple 
Classification Analysis, was not fed into this simple bivariate analysis. A 
correlation between total slave population on the plantation and percentage 
inherited shows, though, that there is a strong negative correlation (-0.9) 



314 



between size and percentage inherited. This is a time-series analysis, which 
usually produces high correlation coefficients. Therefore, the negative cor- 
relation, while undeniable, is probably inflated. 

Even with the caveats of the preceding paragraph in mind, there are several 
points evident here. First, the position of the father influenced the posi- 
tion of the son, as would be expected. Applied only to slaves, conservation 
of property within the family meant that slaves were also protected, since 
ownership of slave property was strongly influenced by systems of inheri- 
tance. Thus, an amazingly high percentage of a testator's slave property 
reappeared in the estates of his heirs. As the heirs prospered, though, the 
impact that the father's slaves had on his son's slave property diminished. 
Patrimony in slaves, so to speak, became the basis of a slave quarter but not 
necessarily the totality of it, since the larger the slave population, the 
more likely it was that there were "foreign" (i.e., slaves obtained from 
sources outside of his own family) slaves in it. It is entirely likely, too, 
that these .slaves were foreign not only to Cordes' properties, but to the 
colony as well. Between 1752 and 1756, slave imports to Charleston grew by 
566.0 percent, and between 1757 and 1762, slave imports to Charleston grew by 
20.0 percent (Bentley 1977:69, 74). Early in the eighteenth century, South 
Carolinians had begun to import slaves directly from Africa, and hence, the 
majority of these imported slaves very likely came from Africa and not from 
one of the other mainland or island colonies (Curtin 1969:145). Finally, kin- 
ship among whites affected the distribution of property, including the black 
population. In this case, inheritance customs contributed to continuity in 
ownership of slaves and hence to the stability of the slave community. 

It is evident that heirs did not sell the slaves that they inherited. In the 
John Cordes Estate Account Book, in fact, there is clear evidence that Samuel 
Cordes deliberately maintained the integrity of John Cordes' slave popula- 
tion. According to the terms of John Cordes' will, the estate was to be kept 
together until the occasion of his daughters' marriages or until his sons 
reached their majority. 

When Elinor Cordes married Theodore Gailliard in 1764, a total of 14 slaves 
were drawn off and turned over to her husband. Three years later, Catharine 
Cordes married Samuel Prioleau, a merchant in Charleston. Prioleau sold the 
13 slaves he obtained by right of his wife back to Samuel Cordes, who acted 
on behalf of the John Cordes estate and divided the Prioleau lot between the 
two remaining heirs, John and Thomas Cordes (John Cordes Estate, Account 
Book, 1756-1798, pp. 8, 30). 

It is equally as likely that Samuel Cordes was concerned with protecting the 
integrity of the estate, as that he repurchased these slaves out of concern 
for bonds among the slaves. In order to explore the question of how con- 
scious whites were of bonds among blacks and how interested they were in 
maintaining and protecting those bonds, relationships among the slaves in 
John Cordes' estate, whose names are known, were investigated. Herbert Gut- 
man, in his critical book, The Black Family in Slavery and Freedom, 
1750-1925 , demonstrated that patterns in slave names are a key to familial 
relationships among slaves, since slaves tended to name their children for 
their kindred, particularly for their male kindred (Gutman 1976:189-190). 
Considering only namesakes in the five lots of slaves selected from John 



315 



Cordes' estate, names from one list that matched with names on the same or 
another list were considered to mark a kinship relation. No effort was made 
to guess the nature of the relationship beyond inferring that the prefix "Lit- 
tle" meant a child, "Young" meant a young or middle-aged adult, and "Old" 
meant a grandfather or grandmother. 

These assumptions obtained the following results. No namesakes were divided 
in the widow's share, withdrawn in 1764. In the Gailliard share, also with- 
drawn in 1764, three slaves, who were evidently prime or elderly, were separ- 
ated from namesakes and two women were kept together. In the Prioleau share, 
separated out in 1767, three slaves were separated from namesakes. When this 
share was divided between John and Thomas Cordes, two, a woman and a child, 
were not reunited with their namesakes, and one, Old Harry, was put back in 
the lot that contained Little Harry. Both Old Harry and Little Harry were 
separated from Big Harry, who was probably the adult, father to Little Harry 
and son to Old Harry. Since John's and Thomas' lots came last, several separ- 
ations had already been made. Five slaves in John's lot had already been 
parted from namesakes, and one new separation was made when drawing off his 
share in 1768. In John Cordes' share, there were at least two sets of name- 
sakes, although one of these sets was a grandfather/grandson relationship. 
Four slaves in Thomas' share were also separated from namesakes, and two 
slaves, Old Culley and Little Culley, possibly grandparent and grandchild, re- 
mained together. 

There is ample evidence in the wills and other plantation records that testa- 
tors were aware of relationships among the slaves. The Ravenels' various 
lists of slaves recorded the births of slave children, noting both father and 
mother. "Henry Ravenel , moreover, organized his slaves according to the plan- 
tation on which they lived and the way in which he had obtained them. Thus, 
he kept the records of the slaves inherited from his father Rene separated 
from other records of slaves (Check Book of Slaves, 1771-1850, Thomas Porcher 
Ravenel, Family Papers, 1731-1906, SCHS). Among the Cordes, references in 
the wills clearly recognize bonds among the slaves. Thomas Cordes (d. 1762) 
left his wife Ann several slaves including Mustee Molley "and her Child Lip- 
peTle." James Cordes (d. 1789), who died without children, left his grand- 
nephew William Cordes the woman Joan and her four children, which he named, 
and Fanny and her two children, which he also named. Samuel Cordes himself 
left his wife two carpenters and the "house servant Martha and her children" 
(Will of Thomas Cordes, Record of Wills, Vol. 10, Book B, 1760-1767, p. 450; 
Will of James Cordes, Record of Wills, Vol. 23, Book B, 1786-1793, p. 414; 
Will of Samuel Cordes, Record of Wills, Vol. 26, Book B, 1793-1800, p. 506, 
SCDAH). 

The tendency to leave slave women and their issue to an individual was one 
means by which women and their young children, at least, tended to stay to- 
gether, facilitating the identification of black women with their children. 
Thomas Cordes, Jr. (d. 1799) left his daughter Elizabeth his seamstress 
Satyrah and her issue (Will of Thomas Cordes, Jr., Record of Wills, Vol. 27, 
Book C, 1793-1800, p. 960). When Elizabeth married in 1819, Satyrah "and her 
issue Dinah and Tony" appeared in the list of slaves attached to her marriage 
settlement (Marriage Settlement of Elizabeth Cordes, Marriage Settlements 8, 
p. 38-41, SCDAH). As the Ravenel records indicate, white masters were also 
aware of relationships between fathers and their families, and in 1799, 



316 



Thomas Cordes, Jr., explicitly bequeathed to his son James "my Driver Mush 
and his wife Dinah and her youngest child" (Will of Thomas Cordes, Jr., Rec- 
ord of Wills, Vol. 27, Book C, 1793-1800, p. 960). 

The slave family is discussed in greater detail in the following section, 
since family is critical to discussion of stability and the internal workings 
of black life. It is important in this context to observe, however, that 
whites clearly recognized black family organization but did not respect it 
entirely when the needs of the estate were considered. The practices which 
exceeded the dictates of a given individual and, in fact, influenced the de- 
cisions he made (e.g., kinship, localism) worked to restrict the damage to 
black bonds that divisions might have made. Thus, white inheritance prac- 
tices tended to stabilize slave ownership within the extended white family, 
although they possibly affected the immediate slave population from planta- 
tion to plantation. Since the extended white family tended to group their 
plantations within a restricted geographic area in the late eighteenth cen- 
tury, dislocation among blacks was not as drastic as it might have been. 
Finally, the pattern of divisions evident in separating out lots in John 
Cordes* estate indicates that although the nuclear family's bonds between 
parent and child were strained, the bond between grandparent and grandchild 
could be preserved when the former was broken. The pattern of division, 
therefore, was conducive to the preservation of an extended black family 
within a restricted area although the nuclear family was attenuated. Allan 
Kulikoff has found a similar pattern of limited dislocation in the seven- 
teenth century and early eighteenth century Chesapeake. By the 1730s, he 
argues, slave kin networks had begun to take shape and over time, "short 
distance sale and estate division spread kin groups over the county" (as 
summarized by Gutman 1976:342). 

The preceding discussions related to the impact primarily of inheritance 
strategies on the slave populations. In the early nineteenth century, Thomas 
Cordes (d. 1806) and his heirs began to sell slaves off. Some of the bills 
of sale have survived, listing different kinds of information about the 
slaves. 

Table 69 summarizes the available information for these six transactions. In 
1790, Thomas Cordes sold two slaves to two Charleston merchants. Eight years 
later, he sold 22 slaves to Catharine Cordes of St. John's, Berkeley, 
probably his mother, who was living in St. John's at that time. In 1800, he 
sold his sister-in-law 15 slaves, under the restriction that she hold them on 
behalf of his wife, and in 1814, Charlotte Cordes sold 28 slaves to Philip 
Porcher on behalf of the estate of Thomas Cordes. Dr. Samuel Cordes, the son 
of Thomas and Charlotte Cordes, who was living in Santee, mortgaged 63 slaves 
to Gibby and Waring, including all slaves known to have been inherited from 
his mother's estate, and probably in 1836, M. Catharine Cordes sold Solomon 
Clarke 18 slaves. 

Four of these six exchanges were either within the parish or involved kin in 
the parish or the two adjoining parishes. The eighteen slaves sold to Clarke 
were probably resident on Yaughan, and thus, ownership followed slave residen- 
cy, since he purchased Yaughan in January 1836. In the case of Dr. Samuel 
Cordes, when he moved to Santee, another adjoining parish, he evidently took 



TABLE 69. Bills of Sale and Mortgage, 1790-1836 



317 



From 



To" 



Identification No. of Slaves 



Year 



1790 Thomas Cordes, of 
St. Stephen's 



Timothy Ford 
William Henry 
DeSaussure 



Merchants of 
Charleston 



1798 


Thomas Cordes, of 
St. Stephen's 


Catharine Cordes 


Mother, St. John's, 
Berkeley 


22 


1800 


Thomas Cordes, of 
St. Stephen's 


Margaret Cantey 


Sister-in-law, 
St. Stephen's 


15 


1814 


Charlotte Cordes for 
Estate of Thomas 
Cordes 


Philip Porcher 


Planter, St. 
Stephen ' s 


28 


1834* 


Dr. Samuel Cordes, 
of St. James, Santee 


Gibby & Waring 


Bank of South 
Carolina 


63 (13) 


[1836] 


M. Catharine Cordes, 
of St. Stephen's 


Solomon Clarke 


Planter, St. 
Stephen's 


18 



* Mortgage of 63 slaves; 13 match with slaves known left to him in Charlotte 

Cordes' Will (Will of Charlotte Cordes, 12 June 1826, recorded 26 May 1827, 

Record of Wills, Vol. 37, 1826-1834, pp. 238-241, SCDAH). 

Sources: Miscellaneous Records, Vol. 000, pp. 270-272, SCDAH. 

Miscellaneous Records, Vol. 50, p. 239, SCDAH. 

Miscellaneous Records, Vol. LLL, p. 40, SCDAH. 

Miscellaneous Records, Vol. ZZ, p. 137, SCDAH. 

Miscellaneous Records, Vol. 5R, p. 147, SCDAH. 

Miscellaneous Records, Vol. 3W, p. 255, SCDAH. 



318 



his share of the estate's slaves with him. Only two slaves were sold away 
from the area and from the circle of kindred/neighbors. Sixty-one (excluding 
the 13 slaves known to have gone from Charlotte Cordes to her son Dr. Samuel 
Cordes) stayed within the familiar circle of kindred and neighbors. This 
early nineteenth century pattern of sale echoes the restricted network of 
exchanges of the mid and late eighteenth century. In this instance, as had 
been the case with inheritance customs, it was conducive to stabilizing the 
pool of slaves who resided in a limited area although the community of slaves 
on a single plantation might suffer periodic, short-distanced dislocation. 

The Black Community 

The preceding parts of this section detailed practices among the whites that 
defined the boundaries of the slaves' world: occupancy, provisioning and ow- 
nerships. Conventions bounding white behavior, namely localism and kinship, 
were found to have profound impact on the slaves' experience. Both height- 
ened the insularity of the parish, which was reinforced by the economic 
autonomy of the plantations. The black majority was relatively insulated 
from the greater white world, and the degree of independence and travel 
blacks did enjoy facilitated communication within the slaves' neighborhood of 
plantations. This softened the impact of sales of slaves within the parish 
or between adjoining parishes and is consistent with a pattern that streng- 
thened, according to Herbert Gutman, an African identification with the wider 
kinship network (Gutman 1976:211-212). The slaves owned by the Cordes family 
in the eighteenth century have not left a written record to confirm or con- 
test Gutman's link between demographic characteristics and the perpetuation 
of African conception of the family. The preceding part of this section has, 
however, shown that the demographic pattern Allan Kulikoff described among 
slaves in the colonial Chesapeake was replicated in this stretch of the low 
country. 

The relationship between the slaves' world and the greater colonial horizon 
is one side of the historical record. The other is the internal structure of 
the black community. Philip Morgan's dissertation is a rich source of demo- 
graphic information and supplies a comparative framework for information de- 
rived from the Cordes family papers and other documents describing St. Ste- 
phen's Parish. Since the Cordes' inventories did not consistently indicate 
the relative age of the persons listed, a reliable age structure of the black 
population could not be obtained. Much of Morgan's analysis was, therefore, 
not repeated in this project. The documents do produce the following cate- 
gories of data: size of plantation, male/female ratios, family size and fre- 
quency of family size. 

Table 70 summarizes information relating to the size of slave populations on 
the plantations owned by members of the Cordes family over time. Morgan 
estimates that the typical eighteenth century South Carolina plantation 
housed from 50 to 62 slaves. In the 1730s, half of the slaves lived on plan- 
tations with 20-50 slaves, and by the 1770s, half of all inventoried slaves 
resided on large plantations. The change that he observed concerned the dis- 
appearance of the relatively small slaveholding operation and the consolida- 
tion of larger holdings over the course of the eighteenth century (Morgan 
1977:1, 4, 7). Ownership relates more to the stratification of the white 
population owning slaves. More important for ascertaining the perspective of 



319 



TABLE 70. Slave Demography in Cordes' Inventories, 

1 745-1 806 

Year Owner Size Men: Women 

1.6:1 

1.1:1 

1.3:1 

1.3:1 



1.3:1 
1.3:1 



(1) 


1745 


Isaac Cordes 


114 


(2) 


1749 


Col . Thomas Cordes 


148 


(3) 


1757 


John Cordes 


53 


(4) 


1764 


Estate of John 
Cordes 


65 


(5) 


1764 


Thomas Cordes 


74 


(6) 


1765 


Henrietta Catharine 
Cordes 


30 


(7) 


1796 


Samuel Cordes* 


408 


(8) 


1796 


Thomas Cordes Jr.** 


82 


(9) 


1796 


Curriboo Plantation 


103 


(10) 


1807 


Thomas Cordes*** 


47 



1.06 
1.05 
1.4 
1.04 



* Represents Samuel Cordes' entire holdings. 
** His slaves at Mil ford. 
*** Thomas Cordes owned Yaughan. 

Sources: Inventory of Isaac Cordes, 9 August 1745, Inventories, 67A, 
1732-1746, pp. 316-332, SCDAH. 

Inventory of Colonel Thomas Cordes, 21 April 1749, Inventories, 
Vol. B, 1748-1751, pp. 141-145, SCDAH. 

Inventory of John Cordes, 22 January 1757, Inventories, Vol. 5, 
1756-1758, pp. 22-28, SCDAH. 

John Cordes Estate, Account Book, 1756-1798, pp. 1-2, CC. 

Inventory of Thomas Cordes, Undated, Inventories, Vol. V, pp. 
492-494, SCDAH. 

Inventory of Henrietta Catharine Cordes, 10 January 1765, Vol. W, 
pp. 221-222, SCDAH. 

John Cordes Estate, Account Book, 1756-1798, pp. 134-146, CC. 

Inventory of Thomas Cordes, 22 June 1807, Inventories, Book D, 
1789-1810, p. 429, SCDAH. 



320 



the slaves is the fact that over the eighteenth century, South Carolina's 
black population became concentrated in a limited area, maximizing blacks' 
opportunities for contact with each other. Thus in 1720, 82 percent of the 
colony's slaves lived in parishes that were over 60 percent black, which 
included St. John's, Berkeley, and 43 percent lived in parishes that were 
over 70 percent black. By 1790, all but two of the lowland parishes were 
over 70 percent black, and these parishes comprised 85 percent of all the 
state's slaves. Black population density was highest in the Santee region, 
which included the Parish of St. Stephen (Morgan 1977:4, 7). Clearly, the 
slaves owned by the Cordes family resided in a demographic setting similar to 
that of the majority of black South Carolinians, although their white masters 
represented an elite segment of the white population (see above, "The Owner- 
ship of Slaves") . 

The continuity in ownership demonstrated earlier indicated that the slave com- 
munities on these plantations were not entirely determined by slaves a slave- 
owner inherited from the parent generation. Rather, growth was probably 
brought about in the 1750s by acquiring slaves who were very possibly recent- 
ly imported from Africa, and through natural increase. The conclusion has 
two significant features. First, the chance for discontinuity as a result of 
sale was small, and second, importation provided re-contact with Africans for 
slaves born in South Carolina. Working from another perspective, Morgan also 
found less likelihood for dislocation among slaves on the nature of South 
Carolina's staple. "In South Carolina," he writes, "the combination of large 
landhol dings and the non-exhaustive nature of the rice crop meant that slaves 
and their descendants had more chance to remain on the same plantation than 
was the case in Virginia" (Morgan 1977:16). 

Table 70 is misleading in that it suggests that slave ownership declined over 
time. Instead, the numbers of slaves owned by Isaac Cordes' son and grandson 
(column (3) and column (10)) declined, but those owned by Colonel Thomas 
Cordes' descendants increased. This had as much to do with a series of 
deaths of men without sons to inherit as it did prudent investment. Samuel 
Cordes, for example, died in 1796 possessed of several plantations and sets 
of slaves because of fortuitous Inheritances from his brothers and uncle. 
His total of 408 slaves was divided among several plantations. The table 
does indicate the range of si avehol dings in the family, and indicates that in 
St. Stephen's, plantations worked by the same family varied in size. Even 
the smallest of these operations, the 47 slaves owned by Thomas Cordes at 
Yaughan in 1806, was well into the category of large plantations. Tra- 
ditionally, the benchmark is 20 slaves, although plantations in South Caro- 
lina tended to be much larger. 

Calculation of the size of the plantation slave communities was a preliminary 
step in calculating the ratio of men to women in the slave community. Except 
for Curriboo in 1796, the ratio tended to equalize over the century. This is 
consistent with Morgan's findings. In the 1730s, the imbalance between men 
and women was greatest, reflecting the preference for African males in the im- 
portation of slaves. This preference was also responsible for the low rate 
of natural increase in the black population. Later decades, however, were 
not as seriously affected, despite the massive influx of Africans in the 
1750s. The 1760s and 1770s, therefore, saw a high rate of fertility that off- 
set the effect of importing slaves (Morgan 1977:187-291). 



321 



Morgan theorizes that the 1740s were probably the critical decade in the 
demographic history of South Carolina blacks, during which the slave popu- 
lation began to grow sufficiently as a result of natural increase to offset 
the effect of the male-dominated importations of the 1750s (Morgan 1977:300). 
By the close of the colonial period, nearly one- third of all of the slave 
children in the colony were found on plantations with equal sex ratios, and 
more equal numbers of men and women in the black population was therefore 
linked to an increase in fertility. Large slave communities tended to have 
more equitable distribution between men and women slaves, and by the end of 
the eighteenth century, in some parishes, almost half of all slaves were 
found on plantations with 100 or more slaves, and only one- sixth of the adult 
slaves lived on plantations with markedly imbalanced sex ratios or with 
members of only one sex. "A large number of South Carolina inventories," 
Morgan concludes from this demographic analysis, "leave no room for doubt 
that many colonial slaves experienced family life" (Morgan 1977:313). The 
structure of slave families was complex, spanning three as well as two ge- 
nerations. In the 1770s, he estimates, some 50 percent of the slaves were 
involved in families, where they were present (Morgan 1977:3.17). The pat- 
terns he has outlined are clearly echoed in the foregoing discussion of the 
Cordes' slaves. 

The Cordes family documents do not lend themselves to reconstructing either 
the age or family structure of the plantation's slave populations. The rec- 
ords of the Palmer family for the late eighteenth century do include lists 
for the distribution of blankets that offer inferences about slave family 
life. The Palmer family, originally spelled Pamor, lived in St. Stephen's 
and had resided in the parish as long as the Cordes family had; in 1754, John 
Pamor, also a descendant of a Huguenot migrant, signed the letter to Alexan- 
der Keith along with Samuel Cordes, inviting Keith to become the pastor for 
the newly created Parish of St. Stephen (Misenhelter 1977:6-7). In a ledger 
spanning the years from 1777 to 1811, an anonymous member of the family noted 
distributing blankets to the slaves six times (Palmer Family, Ledger, 
1777-1809, pp. 1, 2, 3, 167). He gave out the blankets to the heads of the 
slave families, recognizing, in this way, both the family structure among the 
slaves and the person responsible for the social unit. 

Table 71 summarizes data on size of family for four of the six distributions 
for which the information is complete. The tabulations make the following 
assumptions: that the head of household took out as many blankets as he or 
she could, and that the master allowed only one blanket per person; and that 
more than one family may have occupied a cabin. Family size for the period 
as a whole ranged from one to six. In only one instance, December 1796, did 
more than half of the individuals constitute a family of one person, and in 
all cases, a clear majority, close to three-quarters of the population, lived 
in families of two or less. The mean size of family, moreover, hovered 
around 2. This implies that this population was dominated by young adults in 
the process of forming families. The tendency for household size to increase 
over the course of this period 1s consistent with this observation. 

Analysis of the Palmer records refines the impression created by Morgan's sta- 
tistics. It underlines the fact that although the population was more equit- 
ably distributed between men and women at the end of the century than at the 



322 



TABLE 71. Mean Slave Household Size and Distribution of 
Slave Households by Size, 1785-1802 



(a) 


February 1785 




Size 


No. 


% 


Cum. % 


1 


11 


31.4 


31.4 


2 


14 


40.0 


71.4 


3 


6 


17.2 


88.6 


4 


4 


11.4 


100.0 


5 





— 




6 





— 




Total 


: 35 


100.0 


100.0 


mean 


= 2.1 







(b) December 1796 




Size No. 


% 


Cum. % 


1 22 


51.1 


51.1 


2 10 


23.3 


74.4 


3 10 


23.3 


97.7 


4 1 


2.3 


100.0 


5 


— 




6 


— 




Total: 43 


100.0 


100.0 


mean = 1 .8 







(c) February 1799 




Size No. 


% 


Cum. % 


1 17 


40.5 


40.5 


2 14 


33.3 


73.8 


3 5 


11.9 


85.7 


4 5 


11.9 


97.6 


5 1 


2.4 


100.0 


6 


— 


- 


Total: 42 


100.0 


100.0 


mean = 2.0 







(d) November 1802 




Size No. 


% 


Cum. % 


1 17 


39.5 


39.5 


2 16 


37.2 


76.7 


3 4 


9.3 


86.0 


4 2 


4.7 


90.7 


5 3 


7.0 


97.7 


6 1 


2.3 


100.0 


Total: 43 


100.0 


100.0 



mean = 2.1 



Source: Palmer Family, Ledger, 1777-1811, pp. 1 , 2, 3, 167, ISC. 



323 



beginning, the family life produced within the confines of this demographic 
structure was still truncated. Although a significant part of the slave 
population lived in households greater than one, it was mainly homogeneous 
with respect to age. Children were present and evidently brought up in 
households with their parents. Since the generation of the mid and later 
eighteenth century was the last to experience contact with Africans freshly 
imported to South Carolina, these children were in a setting conducive to 
transmitting culture within the household that was still close to its African 
sources. In Chapter III, the changes in the ratio of white to black in St. 
Stephen's in the first decades of the nineteenth century were discussed, 
concluding that by 1850, blacks came into contact with whites far more fre- 
quently than they had in the eighteenth century. Taking the longer view of 
the relationship of the black family to the black community and to the wider 
parish population, both white and black, it is clear that stable black fami- 
lies came to dominate social organization of the South Carolina black popu- 
lation in the decades during which migration from Africa lessened and ended, 
and contact with whites increased. 

Conclusions 

Since the publication of Peter Wood's Black Majority in 1972, historians of 
slavery in colonial South Carolina have been sensitive to the significance of 
blacks in the history of South Carolina. Although Wood emphasized the con- 
tribution of slaves to the provincial culture, most notably their expertise 
in rice cultivation, and the African survivals evident in South Carolina 
slavery, the perspective among most historians has inevitably focused on the 
history of white South Carolinians. Eugene D. Genovese's Roll, Jordan, Roll 
(1974) showed that southern history consists of a dialogue between white and 
black Southerners, but his study has limited relevance for the eighteenth 
century since it is admittedly a study in antebellum culture. Similarily, 
Herbert Gutman's The Black Family in Slavery and Freedom, 1750-1925 (1976), 
although it begins with reference to the colonial period and extends to the 
early twentieth century, applies largely to the mid-nineteenth century. The 
"new" social history of quantitative methods and computer-assisted research 
has supplied sophisticated means of handling the documents and extracting 
information from them. Research by George Terry, Allan Kulikoff and Philip 
Morgan is of this nature, and their conclusions have informed the preceding 
pages. 

As is evident in the citations above, conclusions from scholarly research fa- 
cilitated the interpretation of the documents. All of these studies have 
indicated African survivals in slave life. Morgan, for example, showed how 
African work habits were grafted into the routines demanded by the planta- 
tion, and cited numerous instances of slaves producing items for sale in 
Charleston (Morgan 1977:83-85, 138). Genovese's sensitive treatment of slave 
religion uncovered African survivals in Afro-American Protestantism (Genovese 
1974:280-284). Finally, Wood linked Gullah and Guichee to the underlying de- 
mography of colonial South Carolina, suggesting that their relative position 
in the population and cultural isolation were critical factors in the perpe- 
tuation of Africanisms. 



324 



All of these studies are hampered by the lack of direct testimony. The 
preceding chapters have endeavored to supply direct testimony by looking at a 
form of evidence traditionally alien to the historian. The following chapter 
will synthesize the archaeological evidence. The historical inquiry offers 
the following conclusions to assist in the interpretation of that evidence. 
The plantations were owned by members of the same family, although different 
branches of it, for most of the eighteenth and early nineteenth centuries. 
The men that owned Curriboo were wealthier than those who owned Yaughan. The 
plantations were continuously occupied by whites as well as blacks from the 
mid-eighteenth through the turn of the century, but these planters repre- 
sented an elite stratum within the white population. The plantations were 
settled when St. Stephen's Parish, then called English Santee, was in an 
early stage of development. Very quickly, the parish became the setting for 
the typical slave experience. The demographic evidence is complex, showing a 
high degree of continuity accompanied by imports from Africa in the eight- 
eenth century. The slave population became more stable over the course of 
the eighteenth century in both its ratio of men to women and in its social 
organization. 

It is important to emphasize here that the underlying population was not 
static, but moving toward greater stability from relative instability implied 
by a preponderance of a single sex and age bracket, which is implicit in the 
relative paucity of children characteristic of the early eighteenth century. 
Finally, blacks dominated the countryside, at least in numbers, and the 
plantations in the eighteenth century were relatively isolated from contact 
with Charleston in their daily routines. This situation changed in the 
nineteenth century. As a result of underlying social and economic changes, 
the balance between blacks and whites became more equitable, and the rural 
low country became integrated into an economic system with Charleston as its 
hub. Within the Parish of St. Stephen, the population stagnated and grad- 
ually became denser in the upper portion in the vicinity of Pineville. 



325 



XII. ACCULTURATION AT CURRIBOO AND YAUGHAN 
Introduction 

Questions of who lived where and when are often the total objectives of re- 
search in the archaeology of historic peoples. This report has so far dealt 
with questions of who, where, and when, and also with the question of how the 
inhabitants of the slave quarters lived, within the boundaries imposed by the 
analytical methods used and the resource itself. 

Archaeology rarely deals with questions of the primarily non-material aspects 
of culture. Such questions tend to be closely allied to a type or class of 
artifact or feature even when they are asked. Are figurines an indication of 
cult worship or are they toys (Paddock 1970)? Do the motifs on pottery indi- 
cate that a society was matrilineal or on the verge of collapse (Hill 1970; 
Longacre 1970; Rattray 1966)? Do certain intrusive artifacts show dominance 
or subservience by the intrusive culture or simply trade (Lathrap et al 
1956)? Further, because these questions are usually asked of prehistoric 
sites many competing hypotheses which cannot be satisfactorily proved or dis- 
proved vie for attention. The result is that while such questions are asked 
they usually cannot be answered, primarily because we do not have enough in- 
dependent data on the non-material culture of the group studied or on compara- 
ble groups. Comparing the apparent hierarchical burial practices of the 
Hopewell with such apparent practices among the Maya, for example, can show 
similarities and differences in the types and distributions of artifacts, but 
cannot go beyond this to show the causes for similarities and differences 
since independent variables of non-material culture cannot be held constant. 
Does population pressure or the development of horticulture cause such hier- 
archical systems to develop, or is there something in the belief system quite 
apart from population, technology or environment which causes such apparent 
hierarchical systems? Hypotheses which extrapolate from the material to the 
non-material will remain hypotheses in prehistoric studies without independ- 
ent controls over some classes of non-material variables. 

Historical archaeology offers the opportunity for controlling some of these 
non-material variables. The non-material variables must be controlled so 
that attention can be focused on the archaeological data. Once models can be 
tested in such controlled circumstances perhaps they can be applied more suc- 
cessfully to prehistoric research with a better understanding of the results. 
Lathrap et al (1956:25) noted this problem and stated, 

Comparatively few of the known examples of culture contact 
have been adequately recorded and analyzed. The most impor- 
tant desideratum is the carefully controlled excavation of 
more sites whose histories are known from written records, to 
provide a sound basis for analogical inferences in interpret- 
ing the evidence at fully prehistoric sites. 

The data obtained from the slave quarters at Curriboo and Yaughan lend them- 
selves to a non-material aspect of culture within a reasonably controlled 



326 



situation. There are undoubtedly gaps in the data, variables which are pres- 
ently unknown, and a limited application for the results. In spite of this, 
it is felt that the attempt should be made to examine acculturation at 
Yaughan and Curriboo plantations. This decision was reached after consi- 
dering that no other slave quarters in the Southeast have been so completely 
examined; the historical research has been sufficient to isolate certain 
variables, thus narrowing down the non-material variables to be explained; 
and the data tends to indicate without further elaboration that acculturation 
within the plantations may be a lucrative avenue of study. 

Review of the hypotheses presented at the beginning of this report will show 
that acculturation is implied or explicitly stated in each of them. It is 
clear that throughout the analysis acculturation was considered an important 
aspect of the study. 

Hypothesis 1 stated that Colono was made by and for the slaves, and that one 
variable affecting the hypothesis would be that the diverse backgrounds of 
the slaves would be apparent in the finished pottery. This implies a blend- 
ing, or the results of acculturation, of various cultural backgrounds in a 
single class of artifact within the new slave culture. 

Hypothesis 2 stated that Colono declined in importance and nonlocal ceramics 
increased over time and that this may have been the result of a change in 
status or change over time (acculturation). 

Hypothesis 3 stated that patterns of artifacts are culturally determined and 
will vary from the Anglo-American pattern. It was noted explicitly that the 
magnitude of the differences in patterns would be influenced by the degree of 
acculturation in the slave quarters. 

Hypotheses 4 and 5 stated explicitly that acculturation would be reflected in 
the architecture of the plantations. 

Defining the Problem 

A definition of this acculturation phenomenon must be set forth and criteria 
which can be tested archaeologically must be established to discuss accultura- 
tion at Curriboo and Yaughan. Acculturation was defined by Redfield et al 
(1936:149) as "... those phenomena which result when groups of individuals 
having different cultures come into continuous first hand contact, with sub- 
sequent changes in the original culture patterns of either or both groups." 
Later anthropological literature has been devoted to the acculturation of 
non-material /ideational systems versus material /behavioral systems, the rela- 
tive speed of acculturation of different aspects of culture, and how varying 
degrees of social integration of both societies determine which society 
changes most and how rapidly. A brief perusal of the anthropological liter- 
ature indicates that very little work has been conducted on acculturation of 
more than one cultural group (as was the case among American slaves) or on 
the nature of acculturation when the enslavement of the acculturating group 
is involved (Keesing and Keesing 1971). 



327 



This discussion of acculturation owes much of its direction to Redfield et 
al's (1936) Memorandum for the Study of Acculturation . In this work the 
authors not only defined acculturation, b"ull they IT so summarized those 
aspects of the process which deserved attention by anthropologists (no 
specific mention was made of archaeology). Following their work, accultur- 
ation became popular in the anthropological community especially until the 
mid 1950s. Articles were written about types of acculturation (Freed 1957), 
specific cases of acculturation (Bruner 1956), acculturation of ethnic groups 
in the United States (Spiro 1955), and occasionally on archaeological evi- 
dence of acculturation (Lathrap et al 1956, Hill 1970, Longacre 1970, White 
1975, and Henry 1980). Apparently the subject was felt to be exhausted, or 
more likely ceased being a fad, so that since the 1950s relatively little new 
work has been conducted exclusively on acculturation (Honigman 1973:1104). 

Despite this vast amount of work between the 1930s and 1950s many of the 
avenues of study indicated by Redfield et al have not been entirely explored, 
and certainly from the standpoint of archaeology the study of acculturation 
has barely begun. As noted by Lathrap et al (1956:26), archaeology is in a 
particularly advantageous position to study a subject which relies on time 
depth. Some of the avenues pointed out by Redfield et al (1936:150-152) 
which could be illuminated by archaeological study of eighteenth century 
plantation slavery might be the type of contact involved between groups of 
"markedly different size", of "unequal degrees of complexity in material 
aspects of culture", and where the acculturating group is brought "into con- 
tact with the new culture in a new region." They note situations "where 
elements of culture are forced upon a people" and "where inequality exists 
between groups" on political, economic and social levels. They point to the 
question of what traits will be accepted by acculturating groups and why. 
Finally, they note that the integration of traits requires that the integra- 
tion be viewed over time. 

Acculturation cannot be discussed without touching on status, especially when 
discussing slavery. Low status would be ascribed to a newly arrived African 
slave by both white and slave society if that slave could not speak the com- 
mon language (presumably English), if he/she could not perform required tasks 
properly, and if the individual did not know the leaders in the slave communi- 
ty or how to cope in a new situation. Obviously there were different levels 
of status within the slave community between the newly arrived slave and the 
driver who had acquired considerable responsibility and power over the other 
slaves. The aquisition of such status required acculturation, whether this 
meant acculturating into slave quarter society or into white society. The 
aquisition of status also Implies learning what status is, how to achieve it, 
and actually modifying one's behaviour to gain it. In the following discus- 
sion, therefore, status and acculturation will be discussed together. 

Archaeology and History at Curriboo and Yaughan 

Attempting to determine the presence of acculturation through archaeological 
data is fraught with difficulty. Archaeology must deal with a distorted sam- 
ple of the material culture and material goods evident to the archaeologist 
may not imply acceptance of the non-material trappings normally associated 



328 



with them, despite Lathrap et al's (1956) implied attempt to equate the two. 
These problems, which would be insurmountable in prehistory (see, for exam- 
ple, Meggers 1975), have been partially mitigated by historical research for 
this project. 

The historical research has established constants at the plantations which 
allow certain conclusions to be drawn from the archaeological data. An ini- 
tial problem which would have made it impossible to draw conclusions from the 
archaeological data was the question of the relative demographic stability in 
the slave quarters. Without such stability any conclusions on culture change 
could have been ascribed to "outside" influences rather than internal change 
and acculturation within the plantations. Such stability was present at 
Curriboo and Yaughan, as explained in detail in Chapter XI. Any changes 
apparent in the archaeological data therefore have a high probability of 
being the result of local change rather than the result of wholesale changes 
in the slave population, with totally new groups of slaves replacing the 
previous population and bringing in new ideas from outside the region.. 

Based upon a study of plantation and parish records, the slaves at Yaughan 
were virtually isolated from white culture between the 1740s and the Revolu- 
tionary War. Thereafter their contacts with white culture undoubtedly in- 
creased, although they still outnumbered the white population and remained 
fairly isolated when in the slave quarter. Until at least the Revolutionary 
War any changes in the archaeological record would probably not have been the 
direct result of forced acculturation to the extent that it was later to 
become. 

The historical research has also established that before the decade of the 
Revolutionary War the number of newly imported African slaves may have sig- 
nificantly increased slave contact with non-western cultures. After the 
Revolutionary War such contact decreased significantly, so that the slaves at 
Yaughan and Curriboo were essentially isolated from any direct contact with 
African cultures. Archaeological ly, this would mean that more evidence of 
African culture should be evident prior to the 1780s and that different 
cultural patterns should be evident after that time, as the slave quarters 
evolved their own cultural patterns based upon African traditions and upon 
more and continuing Anglo-American contact. Although there is a suggestion 
that perhaps there were one or two Indian slaves at Yaughan, historical re- 
search concluded that their impact on the slave culture would have been 
totally obscured by the overwhelming number of African slaves. Further, the 
slaves apparently had the freedom to visit other plantations in the parish, 
share ideas with other slaves in relatively the same conditions as them- 
selves, and to trade goods among themselves and with their masters. This 
would indicate that the archaeological record may also reflect slave accul- 
turation within the parish and not only on Yaughan and Curriboo plantations. 

There are several points on which the historical analysis and the archaeologi- 
cal analysis have independently reached similar conclusions. Such agreement 
further reinforces any conclusions drawn by the archaeology or history alone, 
and provides data upon which any conclusions must be based. 1) History has 
shown that the plantations were established in the 1740s and that Yaughan 
continued in operation until the second decade of the nineteenth century. 



329 



Curriboo continued in operation until around the turn of the eighteenth cen- 
tury. Mean ceramic dates agree with these date ranges. 2) The historical 
research indicates that Yaughan plantation had approximately 20 to 30 slaves 
by the 1750s and increased this to approximately 80 until the late 1790s, 
when the number began to decrease again to around 40 to 50. The archeology 
at Yaughan, where the entire early (pre-Revolutionary War) slave quarter was 
excavated, contained nine houses which probably housed slaves. Assuming 3 to 
4 slaves per house or 50 square feet of floor space per slave (Morgan 
1977:47-48) provides a total of 27 to 36 slaves before the Revolutionary War. 
A second and later slave quarter (38BK75) overlapped the ending occupation 
dates of the earlier quarter, and the earlier quarter was later abandoned, 
leaving only the latest quarter. This agrees with the cycle of slave owner- 
ship noted in the historic documents. 3) Historical documents indicate that 
Samuel Cordes of Curriboo was the more successful manager and better off fi- 
nancially. If the conditions of slave life are an indication of the wealth 
of the owner, then the different financial conditions of the owners noted in 
the history was also reflected in the archaeology. 4) The historical docu- 
ments indicate that Thomas Cordes did not live in residence until the Revolu- 
tionary War. A surface survey of the suspected main house at 38BK75 Locus B 
(see Chapter IV) produced a similar result, indicating that it was inhabited 
well after the initial slave occupation and between the Revolutionary War and 
the first decades of the nineteenth century. 

With these congruences between the historical and archaeological records, and 
the further conclusions drawn from historical analysis concerning the non- 
material conditions of slavery at Yaughan and Curriboo, a unique opportunity 
presents itself to archaeologically examine the acculturation and status of 
plantation slaves in South Carolina from the mid-eighteenth to the early nine- 
teenth centuries. 

Criteria for Acculturation 

Archaeological criteria for acculturation are necessarily material criteria. 
These may hold implications for the non-material aspects of behavior, but the 
criteria must be on the material level. These archaeological criteria can be 
grouped into various classes. 

It has been noted by Keesing and Keesing (1971:353-354) that certain of these 
classes are more easily changed than others and that some classes may indi- 
cate a more superficial degree of acculturation than others. At one end of 
the scale would be the acceptance of an artifact or artifact type by a group 
of people where the artifact fills a preexisting functional niche in the cul- 
ture. Such a situation would be very superficial evidence of acculturation. 
On the other end of the scale would be the acceptance of a wholly new world 
view with complete rejection of the original world view and all or most of 
its material and non-material associations. For these reasons the appearance 
of single artifacts or minor artifact types are given less weight in the 
argument for acculturation. 

Of more significance are changes in two or more types or whole classes of 
artifacts, or, in other words, changes in artifact patterns as used by South 
(1977a). While the appearance of single artifacts or minor artifact types 



330 



might be the result of one or two individuals, the change in the relative 
proportions of whole classes of artifacts as reflected in artifact patterns 
would be much more conclusive proof of culture change. However, artifacts 
may not be as conclusive evidence as settlement pattern changes. 

Settlement patterns change more slowly than artifacts or artifact patterns. 
First, the work and planning involved in changing a house type or settlement 
plan is usually much greater than a change in form or function of an artifact 
type. Second, houses and large structures are used for longer periods of 
time than individual artifacts. That is, the replacement of artifact attri- 
butes and functions is easier than it is for structures since there are more 
opportunities to make changes. Since settlement patterns tend to be more 
conservative and involve more planning than artifact types, changes in house 
type and settlement plan are considered here to be stronger evidence of ac- 
culturation than changes in artifacts. 

Artifact changes, artifact pattern changes and settlement pattern changes in 
an enforced slave setting are suspect because of the fact that acceptance by 
the slaves may have been enforced from outside. From historical research 
this uncertainty about the quality of the evidence may not be so important at 
Curriboo and Yaughan, however, because it has been established that the 
slaves had a certain degree of freedom in such matters. Nevertheless, proof 
of acculturation based on artifacts and houses cannot be as conclusive as 
changes in less obvious (to the slave masters, at least) aspects of culture. 
This is the same argument used by ceramic typologists when choosing ceramic 
attributes for analysis. The less visible a cultural trait, the more it 
reflects deep-seated attitudes and psychology and the less liable it is to 
conscious change. For this reason those traits not readily accessible to the 
slave master may hold the best evidence for acculturation in the slave quar- 
ter. One such trait is foodways. This does not mean the kinds of food 
eaten, since these were often supplied by the master, as much as it does how 
the foods were prepared and eaten, or the process of eating. Changes in 
foodways in a slave setting where there was no central kitchen allowed a 
certain leeway to slave adaptation that is not found in more readily visible 
and controllable traits such as clothing, tools, and housing. Changes in 
foodways are, therefore, considered to be among the strongest evidences of 
acculturation. 

The archaeological criteria for acculturation, then, are the acceptance of 
new artifacts, new artifact patterns, new settlement patterns and new food- 
ways during the occupation of the sites. If these things change, becoming 
more like the dominant white culture, between the beginning of the occupation 
at the sites in the 1740s their abandonment in the nineteenth century, we 
will be able to conclude that acculturation has taken place. The magnitude 
of the changes and where and when they occur are the subject of the following 
sections. 

Acculturation Periods 

From approximately 1745 to 1780 the slaves at Yaughan were gradually exposed 
to Anglo-American cultural influences. From a part-time overseer and perhaps 
even part-time habitation by the slaves themselves to a full time overseer 
and the permanent habitation of 38BK76, slave life at Yaughan settled into a 



331 



routine based on indigo and rice production. This life was characterized by 
its isolation from Anglo-American society, especially on the plantation but 
also within the parish generally, by the slow increase in the size of the 
slave population from natural Increase and the influx of new slaves from 
Africa, and most importantly by a general stability and continuity in the 
slave community. 

From approximately 1780 until about 1805 the size of the slave population at 
Yaughan doubled. During this time the new slave quarter at 38BK75 was built 
and at the end of this period the old slave quarter at 38BK76 declined until 
the major portion of the latter site was no longer used by the late 1790s. 
This period is characterized by a dramatic increase in the number of slaves 
of non-African birth, stability within the corps of the Yaughan slave popu- 
lation, many more contacts with Anglo-American society, including a resident 
owner, his growing family and a growing white population in the parish, and 
by a general loss of the relative isolation of previous years. Such an in- 
crease in the slave population and contacts with Anglo-American culture must 
have resulted in more regimentation in the organization of the work force and 
in more restrictions on the freedom of slaves to maintain visible signs of 
African speech patterns, patterns of dress, and forms of acting when in the 
presence of whites. Since there appears to have been an intensive effort by 
Thomas Cordes at Yaughan to pursue indigo and rice, requiring a doubling of 
his slave force, it is also likely that the slaves had less free time to 
pursue indigenous crafts and growing or foraging for their own food. The 
lack of indigenous craft and food resources, in turn, would have needed to be 
filled by outside sources based upon slave needs, but probably supplied by 
the owner. This would have greatly influenced the types of food and material 
items found archaeologically. 

From approximately 1805 to 1825 Yaughan Plantation saw a decrease in the 
number of slaves and the uses to which they were put. Thomas Cordes' widow, 
Charlotte, appears to have moved to Pineville during this period and began 
renting out her slaves to other plantation owners. This period is charac- 
terized by the maintenance of a corps of slaves descendant from the original 
Cordes slaves, thereby maintaining a stability and continuity in the slave 
quarter. However, acculturation was undoubtedly affected by slaves who had 
been rented out and returned to the plantation with new experiences and 
perhaps more profound changes in their cultural outlook than would have 
resulted in earlier periods from short visits to friends on other planta- 
tions. From an historical point of view, therefore, one should expect the 
archaeology at Yaughan to express more Africanisms during the early period 
(before the Revolutionary War), and a decrease in Africanisms with a con- 
current increase in acculturation during the last two periods. Site 38BK76 
generally represents the first period with overlap into the second period, 
and site 38BK75 represents the second and third periods at Yaughan. 

There is, unfortunately, no site which only represents one period except for 
38BK245, which represents the Curriboo equivalent to the second period at 
Yaughan. The slave quarter at Curriboo operated from approximately 1740 to 
1800, or during the time span of periods one and two at Yaughan. During this 
time the owner was generally in residence at least part of the year and ap- 
pears to have taken more interest in and to have done a better job of running 



332 



his plantation profitably. The slave population, although larger than that 
at Yaughan, exhibited the stability and continuity evident at Yaughan. 
Curriboo produced naval stores and bricks as well as labor intensive agricul- 
tural products and toward the end of its occupation, contained what appears 
to be an office or administrative building which indicated the seriousness 
with which Samuel Cordes conducted his affairs. A situation in which the 
slave population is relatively large, the owner is in residence at least on a 
part-time basis, and the plantation was profitably run and maintained would 
be more similar to the second period at Yaughan than to either the first or 
third periods. 

Arti factual Evidence 

Certain categories of artifacts, artifact patterns, architecture and foodways 
at the plantations were analyzed to determine if such shifts in acculturation 
were operating in the archaeology. Archaeological evidence that the slaves 
had accepted certain Anglo-American cultural items prior to the Revolutionary 
War is shown by artifacts such as an iron which may have been used to iron 
slave or the overseer's clothing, a thimble for sewing clothing or decorating 
Colono pottery, a twisted lead pencil, indicating that the slaves were ex- 
posed to the concept of reading and writing even though they may not have 
been legally allowed to learn these skills, and a mirror probably for per- 
sonal use by a slave. These artifacts do not show conclusively that the 
slaves had accepted the non-material cultural patterns associated with such 
items by Anglo-American society, but at least they were exposed to such 
items. It should also be noted that the artifacts at 38BK76 represented a 
small fraction of the 22,666 total artifacts recovered from the site. At 
38BK245 there were virtually no such individual items clearly attributable to 
the slave occupation. At 38BK75, hypothesized to be the most acculturated 
site, there was a coin indicating that at least some slaves were aware of a 
money economy and probably took part in it, a finger ring, and umbrella 
struts. While these Anglo-American artifacts again show acceptance of 
material goods, their proportion of the 6689 artifacts at 38BK75 is yery 
small and inconclusive. 

Perhaps more convincing evidence for and the comparative degree of accultur- 
ation is found in minor artifact groups which appear at two or more sites. 
Normally, conclusions drawn from low percentages of artifact groups can be 
explained by sampling error, i.e., small sample size or the method by which 
the sample was selected or retrieved. Although sampling error cannot be 
completely ruled out, it should be remembered that entire structures and 
their associated features were excavated at each site resulting in nearly 
total artifact retrieval. Also, large areas of all three sites were in- 
tensively collected, and all large features, even those not clearly as- 
sociated with structures, were excavated. Indeed, virtually the entire slave 
quarter at 38BK76 was exposed. The sample size at all three sites is sub- 
stantial, negating, to a certain extent, any non-randomness inherent in the 
field methods. While the following conclusions may be tentative, they are 
certainly suggestive and more solidly based than those drawn from sampling 
parts of a few structures or areas within a site. 



333 



The largest single minor artifact type was glassware. A comparison of the 
glass sherds in Table 72 shows that as a percentage of total artifacts, 
stemmed glasses, tumblers, and pressed glass represented .19 percent at 
38BK75, whereas at 38BK76 and 38BK245 they represented .07 percent or two and 
one half times less than at 38BK75. Since the percentages at 38BK76 and 
38BK245 were the same and 38BK76 had more, this may indicate that glassware 
is more representative of the third period at Yaughan (1805-1825). Defini- 
tive archaeological dating at such a level was not possible, however. 



Table 72. Artifact Types Compared 
as Percentages of Total Artifacts 



38BK75 38BK76 



38BK245 



Nonessential 








Gl assware 


0.19 


0.07 


0.07 


Tableware 


0.10 


0.03 


0.12 


Non-ceramic 








Kitchen Group 


8.6 


8.8 


11.7 


Colono Ceramics 


38.1 


66.4 


55.8 


Nonlocal Ceramics 


15.3 


7.2 


7.5 


Catawba Ceramics 


4.4 


0.62 


0.29 


All Ceramics 


57.7 


74.2 


63.6 


Architecture 








Group 


23.5 


11.6 


16.2 


Furniture Group 


0.07 


0.05 


0.02 


Arms Group 


0.16 


0.02 


0.29 


Glass Beads 


0.01 


0.09 


0.05 


Tobacco Group 


2.7 


3.3 


5.2 


Fishing Gear 


0.0 


0.02 


0.0 



Gun parts noted on Table 72 made up .16 percent of the 38BK75 artifacts, .29 
percent of the artifacts at 38BK245, and only .02 percent of the artifacts at 
38BK76. If the ownership or use of guns for hunting or chasing pests is an 
indication of acculturation on the part of the slave entrusted with a gun, 
the slaves at 38BK75 and 38BK245 appear to have had the most trust and been 
the most acculturated. The difference between 38BK75 and 38BK245, which runs 
counter to the expectation that time strictly correlates with acculturation, 
may actually be a reflection of the fact that during the third phase at 
38BK75, the number of slaves rented out, and, therefore, not using guns, was 
enough to lower the percentage of gun parts. 

If it is true that slaves generally had less time and freedom during the 
second and third periods to forage for their own food, the time taken up by 
hunting (one of the possible uses of firearms) would have decreased the 
already restricted time allowed for other such activities as fishing and 
plant gathering. Archaeologically, one should see a decrease in the evidence 
for such activities. At Site 38BK76, the total number of fish hooks, possi- 
ble fish hooks, and frog gigs was five and there was no fishing gear at ei- 
ther 38BK75 or 38BK245. This pattern may be the result of sampling at 38BK75 



334 



and 38BK245, but does suggest that fishing was more important at 38BK76. 
Such a low amount of fishing gear is, like the other artifact types so far 
mentioned, inconclusive evidence of acculturation, but the differences are 
suggestive. 

A fourth minor artifact type was glass beads. The ownership and use of such 
beads has tended to be understood as part of the slave artifact assemblage, 
and not part of the Anglo-American cultural pattern. For this reason one 
should expect the earliest period to have the most beads and the latest to 
have the least. The beads at 38BK76 made up .09 percent of the total arti- 
facts. The percentages at 38BK75 and 38BK245 were .01 percent and .05 per- 
cent, respectively. On an acculturation scale from no beads to all beads, 
Site 38BK75 is clearly the most acculturated, 38BK245 the next most accul- 
turated, and 38BK76 the least acculturated. 

Tableware was a fifth minor artifact type which may show acculturation. 
Tableware (knives and forks) 1s .10 percent at 38BK75, and .03 percent and 
.12 percent at 38BK76 and 38BK245, respectively. Here again, sites 38BK75 
and 38BK245 appear to be the most acculturated. This data by itself may mean 
very little, however as will be discussed at the end of this chapter, other 
aspects of foodways also support acculturation. 

A final minor artifact category was furniture artifacts. There is ample 
historical evidence that slaves met many of their own needs in this regard 
(Otto 1975:375-380). An increase in the use of manufactured furniture 
artifacts would tend to indicate that the slaves not only accepted such 
items, but that they either could not or would not make all such items them- 
selves. Site 38BK75 did, indeed, have the most furniture hardware, .07 
percent of the total artifacts. Site 38BK76 had .05 percent and 38BK245 had 
.02 percent. These percentages are ^ery close, but again 38BK75 and 38BK245 
had relatively more such items than 38BK76. This may be because most of the 
furniture material from 38BK76 was found at Structure 76B, the latest struc- 
ture at the site, and also because Structure 76B has two structures, one of 
which was a posthole structure characteristic of later house types. In any 
case, the results of such small percentages are inconclusive, although fur- 
niture artifacts may become a better indicator of acculturation on nineteenth 
century sites when more such material would be available for purchase. 

With this brief discussion, there is a suggestion that individual artifacts 
and minor artifact types reflect acculturation in the slave quarter. It has 
also been pointed out that such evidence is inconclusive since acceptance of 
a single artifact type or minor group does not necessarily indicate accept- 
ance of new habits or outlooks. 

The following artifact type has already been discussed in some detail. It 
was the most numerous artifact at any site and shows more than simple accept- 
ance of a single new item. Colono ceramics required planning, creativity, 
observation, and extensive knowledge of technique and local resources to be 
manufactured. Some of the data collected on Colono will be discussed below 
with foodways. It is unfortunate that details of form could not be studied 
because it might have been anticipated from a detailed study of form that an 



335 



accumulation of small changes would indicate a shift to discernible Euro- 
American stylistic attributes. General form categories that could be 
discerned, however, were discussed in Chapter VIII. Certain conclusions have 
been drawn from the analysis of the Colono at Curriboo and Yaughan, and 
observation of other colonoware collections which support the hypothesis for 
acculturation in the slave quarters. 

It was shown that the manufacture of Colono probably followed the West 
African tradition of open air firing, that Colono was generally not made with 
the coil technique as Indian pottery, that local clays were used, and that 
the finished product was probably not a readily saleable item to non-slaves. 
A chain of data was presented which tends to link Colono with colonowares in 
other regions of the east coast, and to the Caribbean Islands and possibly to 
Africa. Colono and colonoware may be the most African "Africanism" to appear 
on slave sites and as such the single most useful artifact for studying slave 
acculturation. Handler and Lange (1978:144) note that as Barbadian slaves 
became more acculturated the African attributes in their pottery disappeared. 
This change occurred more rapidly on Barbados, where whites outnumbered 
slaves, than on Jamaica or apparently on the mainland. 

The site hypothesized to be the least acculturated (38BK76) had the highest 
percentage of Colono, 66.4% of the total artifacts. Site 38BK245 which 
should show more acculturation, had fewer Colono sherds, 55.8%. Finally, 
Site 38BK75, which is hypothesized to be the most acculturated, had 38.0% 
Colono. As noted in Chapter VIII this seriation does not agree with a strict 
seriation by time since 38BK245 is known from historic documents to be 
earlier than 38BK76. Therefore, a slow natural decrease of Colono based 
strictly on time is incorrect and the changes in the relative amounts of 
Colono must be explained by a different process. 

Status was mentioned as a possible cause for differing amounts of Colono in 
Chapter VIII. This meant that the slaves at Curriboo were materially better 
off because their owner was more successful financially; this trickled down 
to his slaves and was evident in the amount of Colono they felt forced to 
produce to fulfill their needs for cooking and serving. Since the slaves 
apparently had the choice to produce or not produce Colono, they produced 
less and replaced it, not with nonlocal ceramics as might be expected, but 
with other goods. This is shown by the fact that at 38BK76, 74.2% of the 
artifacts were ceramics, whereas at 38BK245 this dropped to 63.6%, or a 
difference of 10.6%. At the same time the ratio of nonlocal ceramics to 
Colono remained nearly the same, .11:1 at 38BK76 and .13:1 at 38BK245. If 
the slaves at Curriboo were replacing the Colono with nonlocal ceramics, the 
total percentage of ceramics should have stayed the same at both sites, while 
nonlocal ceramics, relative to Colono, should have increased as it did at 
38BK75 where the nonlocal to Colono ratio was .40:1. 

What replaced the 10.6 percent decrease in Colono at 38BK245? If not nonlo- 
cal ceramics, is there another category of artifact that fills the void? The 
preceding discussion may give part of the answer. The following figures were 
derived by subtracting the percentages of 38BK76 artifacts from the percent- 
ages of 38BK245 artifacts. Approximately an extra .27 percent of the Curri- 
boo slave artifact assemblage was directed toward gunparts. In this case 



336 



.02% (gunparts at 38BK76) was subtracted from .29% (gunparts at 38BK245), 
giving a .27% remainder. About 1.9% more of the Curriboo artifacts were 
directed toward tobacco pipes (5.2% at 38BK245 minus 3.3% at 38BK76). An 
extra 4.6% was directed toward Architecture Group artifacts (16.2 at 38BK245 
minus 11.7% at 38BK76) and an extra 2.9% may have come from other nonceramic 
artifacts in the Kitchen Group (11.7% at 38BK245 minus 8.8% at 38BK76). If 
these percentages are added together — .27% for gunparts, 1.9% for tobacco 
pipes, 4.6% for architecture, and 2.9% for non-ceramic kitchen artifacts — 
the total is 9.67%, or nearly the amount of artifacts needed to replace the 
low percentage of Colono ceramics. 

Were these expenditures on the slave artifact assemblage necessary to main- 
tain life and health at Curriboo or were they comparative nonessentials, and 
do the apparent choices on the part of the owner or slaves represent accul- 
turation? The answer to the first question must be that the items were more 
or less evenly divided between essentials and nonessentials. However, the 
use of tobacco pipes, certainly a nonessential by Anglo-American standards, 
may show acculturation. As noted previously, guns imply trust, and accul- 
turation is needed for that trust. More architecture items and glassware, 
tableware and other non-ceramic kitchen artifacts also seem to indicate 
acculturation, even if the owner made all of the decisions and the slaves had 
absolutely no choice, which is highly improbable. 

The case for the relative amounts of Colono representing acculturation is 
perhaps clearer when 38BK75 is considered. At that site Colono made up only 
38.1% of the total artifacts, and all ceramics at the site were only 57.7% of 
the total assemblage. The ratio of nonlocal to Colono is .40:1. So, not 
only were ceramics decreasing relative to other artifacts at 38BK75, but 
Colono was also decreasing significantly relative to nonlocal ceramics. 
Clearly, other artifact types and even nonlocal ceramics were replacing 
Colono at 38BK75 in much greater proportions than at Curriboo. The dif- 
ference in Colono at 38BK75 and 38BK76 is 28.3%, whereas between 38BK245 and 
38BK76 it was only 10.6%. This greater difference between the sites was made 
up for primarily by an 11.9% increase in Architecture Group artifacts and a 
8.1% increase in nonlocal ceramics and a 3.8% increase in Catawba ceramics. 
Tobacco pipes actually decreased at 38BK75, and the non-ceramic kitchen 
artifacts remained the same. Variation in the remaining artifact groups is 
too complex to unravel and the percentages involved are less than 1%. 

The overall artifact patterns at the sites following South (1977a) were 
presented in Chapter IX. The patterns of each site were compared to a re- 
vised Carolina Artifact Pattern (CAP). The justification for the revisions 
is presented in Chapter IX, with the Carolina Slave Artifact Pattern based 
upon Yaughan, Curriboo and Spiers Landing. The overall high percentages of 
Kitchen, Architecture, and Tobacco Group artifacts differentiate the CSAP 
from the revised CAP. These overall patterns show a trend from 38BK76 to 
38BK245 to Spiers Landing to 38BK75. This trend involves a decrease in the 
amount of Kitchen Group artifacts and an increase in Architecture Group arti- 
facts. This trend, if continued, would eventually resemble the revised CAP. 
We concluded that the trend showed acculturation of slaves to a hypothetical 
point where the slave pattern would be indistinguishable from a poor Euro- 
American pattern. 



337 



Further, it was noted that the patterns of MCDs at 38BK75 and 38BK76 did not 
follow the normal time lag pattern usually associated with historic sites on 
the economic and geographic margins of society (Adams and Gaw 1977). Site 
38BK75 showed time lag of approximately 15 years. Such a great amount of 
time lag is easily explained if the datable ceramics represent items discard- 
ed by the owner after a normal use-life and then reused by the slaves for an 
additional use-life. Site 38BK76, however, did not show time lag and in fact 
showed the reverse. At first glance this seems to indicate that slaves were 
acquiring nonlocal ceramics before they were manufactured or as new items and 
not as owner discards. The first possibility is obviously false, and the 
second would indicate easier access to Anglo-American material goods and 
therefore faster acculturation at 38BK76 than at 38BK75. This is contrary to 
all expectations on slave acculturation. As discussed in Chapter IX the 
reverse time lag evident at 38BK76 can be explained by examining the relative 
amounts of Colono and nonlocal types upon which the MCD is based. Since it 
was hypothesized and has so far been supported by the historic data that the 
slaves at 38BK76 were relatively isolated until the Revolutionary War and 
then exposed to greater Anglo-American contacts thereafter, it seems reason- 
able to conclude that the majority of nonlocal ceramics date from the post- 
Revolutionary War period (1780-1805) and therefore would give an MCD more 
closely aligned with this second period. If 38BK76 was inhabited from 1780 
to 1795 as has been stated above, an MCD of 1773.0 would agree closely with a 
mean occupation date of 1787.5, assuming there were 14.5 years of time lag. 
Time lag at 38BK75 was 15 years or essentially the same. Interpreted in this 
way, the apparently false reverse time lag actually supports the contention 
that the early period at 38BK76 represents a time when the slaves were much 
less acculturated and more isolated than they were later to become. 

To summarize the data on artifact types and patterns presented so far, it was 
assumed that Colono ceramics represented the least acculturated artifact type 
and as such its presence most clearly reflected the state of acculturation of 
the users. The amount of Colono was very high at the site expected to be the 
least acculturated from historical research. The amount of Colono decreased 
until it reached the lowest percentage at the site expected to be the most 
acculturated. Comparing the more acculturated sites with a base site, in 
this case 38BK76, it was noted that at 388K245 Colono was replaced by essen- 
tials (e.g., Architecture, Gun and some of the Kitchen Group artifacts) and 
by nonessentials (e.g., tobacco pipes and the other Kitchen Group artifacts). 
At 38BK75 the Colono was replaced almost entirely by essentials (e.g., archi- 
tecture artifacts, Catawba ceramics, nonlocal ceramics, etc.). Only a few 
nonessentials, such as stemmed glasses and tumblers 1n the Kitchen Group, may 
have replaced Colono. Furthermore, at 38BK75 some nonessentials noted at 
38BK75 decreased (e.g., tobacco pipes). The total artifact patterns and MCDs 
at Yaughan further supported our conclusions that the slaves were undergoing 
acculturation during the occupation of Yaughan and Curriboo. The artifact 
patterns, especially Kitchen and Architecture Groups, may offer a scale of 
acculturation for other sites as well. 

Some further conclusions can be drawn based on the artifacts. It seems ap- 
parent that acculturation was more complete at 38BK75 than at 38BK245; how- 
ever, the character of the acculturation seems also to have been different at 
the two sites. While replacement types at 38BK245 included fairly even 



338 



amounts of essentials and nonessentials, at 38BK75 the emphasis was on es- 
sentials. The differences in nonessentials and essentials may reflect 
overall economic conditions at the plantations. There is also an aspect of 
the acculturation process lightly touched upon before, that the response to 
acculturation may involve actual choices on the part of the slaves, and that 
the changes seen in the archaeology are not solely decisions by the owner. 

The variation at all three sites in nonessential items, where two of the 
sites were on the same plantation with the same owner and where one owner 
controlled all three sites at one point, indicates that the slaves were prob- 
ably making at least some of the choices or communicating their perceived 
needs and wants to the owner who then acted upon them. This is important 
because evidence of voluntary acculturation presents a much more satisfactory 
and stronger case for non-material acculturation than the enforced acceptance 
of objects. 

Architectural Evidence 

Architectural evidence for change within the slave quarters at Yaughan and 
Curriboo may support the acculturation hypothesis. The basic data for this 
brief summary of architectural evidence is given in Chapter VII. The con- 
struction techniques at Curriboo and Yaughan included several types: trench 
foundations with fairly closely set posts, leading to a conclusion that the 
superstructure was mudwall with a lightweight thatched roof; trench founda- 
tions with more widely spaced and occasionally prepared posts, leading to a 
conclusion that the superstructure was frame; posthole foundations with 
widely set posts, also leading to a conclusion that the superstructure was 
frame; and a brick foundation of widely set piers with a brick chimney, 
leading to a conclusion of a frame superstructure. There is occasionally 
evidence for interior walls or partitions in the larger structures and for 
tables, benches, beds or other Interior furniture in the smaller structures. 

At Yaughan there is a sequence of structures from poorly or unevenly con- 
structed trench structures to better constructed posthole structures. At 
Curriboo there is a sequence from well built trench structures to the brick 
piered structure. Within each plantation evolution of architectural tech- 
niques is evident. Between plantations there is evidence that the poorly 
built mudwalled trench structures at 38BK76 are coeval with the well built, 
usually frame, trench structures at 38BK245. The orientation, alignment and 
care with which the Curriboo structures were built as opposed to the early 
structures at Yaughan leads to a conclusion that there was probably more 
control over the slaves at Curriboo than at early Yaughan. This reinforces 
the historic documents and the previous statement that Curriboo is more 
similar to the second period at Yaughan than to the first or third periods. 
Unfortunately, neither of the owners' houses could be examined to determine 
whether the sequence noted in the slave quarters approached an upper class 
Anglo-American pattern of the same era. However, there is one structural 
complex at 388K76 which appears to be outside the slave construction se- 
quence. Its artifacts, associated features and complexity have resulted 1n 
its tentative identification as the overseer's house or complex. If this was 
the overseer's house it represents an example of Anglo-American structures of 



339 



the period and does indeed point the direction for future structures at 
Yaughan. Its post distance and hypothesized superstructure resemble the 
later structures at 388K75 more than the contemporary or sometimes later 
structures at 38BK76. The construction sequence and the possible overseer's 
complex indicate that change occurred in slave architecture through time and 
probably approached a more Anglo-American pattern. 

Further support that the slaves were approaching a more Anglo-American pat- 
tern are the width to length ratios at the sites and in the comparative 
literature. The earlier domestic structures where slaves have already been 
shown to have had more freedom of action and to have been less acculturated 
are more rectangular. In fact a seriation based on W/L ratios places the 
sites from 38BK76 Group B and 38BK245 to 38BK76 Group B to 38BK75, with 
38BK75 being the most square and the closest to the Euro-American structures 
in the literature. 

We conclude that the architecture supports a sequence of acculturation simi- 
lar to that observed in the artifacts and artifact patterns and hypothesized 
from the historical research. Such subtle changes as W/L ratios over an 
extended period was most likely not the result of direct imposition by an 
outside force. A change imposed from outside would have been more drastic 
and clearcut. One day houses would have been built in one fashion, and the 
next they would have been clearly different. A subtle change implies an 
unconscious acceptance of new views on what a house should look like and how 
it should be built. The variation noted in the earlier slave quarter at 
Yaughan seems to indicate that there was a general consensus that houses 
should be rectangular, but exactly in what proportions was not generally 
agreed upon or considered important. At Curriboo the rectangular! ty becomes 
more standardized and apparently more important. At 38BK75 the move to 
squareness has clearly begun with the standardization of size as well. 

Food Preparation 

The last subject for consideration of acculturation presented here is that of 
foodways. This subject has been presented in some detail in Chapter X with 
respect to food sources available to and used by the slaves. The processes 
of food consumption have been briefly touched upon throughout the report. 
The following paragraphs will bring our conclusions from individual sources 
together and develop overall conclusions on how foodways show acculturation. 
It has been stated that the process of food procurement and preparation 
should show the subtle non-material aspects of acculturation more clearly 
than examination of the artifacts as objects. This is so because processes 
usually are not as obvious to outsiders and therefore not as liable to con- 
trol by outsiders. It is possible that slaves, or any accul turating group, 
would outwardly accept certain objects or even houses and occupations if they 
are forced to, without accepting the changes in mental outlook that these 
seem to Imply. A good example of this is quoted by Bruner (1956:612) as 
spoken by an unacculturated Mandan-Hidatsa, "We Indians are just like those 
lizards that change their colors. When we are with whites we act one way and 
when we are with Indians we act another." Foodways is one activity that was 
conducted when slaves were with slaves. 



340 



It has been established that the slaves at Yaughan probably had a primarily 
vegetarian diet of grains. It has been suggested that this diet was sup- 
plemented by other plant foods which would not have been preserved. Meat 
sources are nearly absent, although fishing gear suggests that fish may have 
been part of the diet and not have been preserved. Large mammals, though, 
are virtually nonexistent either in the earlier or later periods. At 
Curriboo the diet appears to have been more mixed, with much more faunal 
material despite poorer preservation conditions than Yaughan. Most of the 
foods at both plantations whether plant or animal, were domestics. Since few 
slaves were obtained from parts of Africa which were nonagricultural , this 
diet would not have been wholly unusual for new arrivals. In any case, they 
may not have had much choice in what they ate. 

On the subject of food processing certain conclusions can be stated on the 
basis of the archaeology: 

-Slaves cooked on-site. There is ample evidence for cooking on-site present- 
ed by fire blackened Colono pots, some with burned organic residue. 

-Slaves cooked out-of-doors. There is evidence for one certain and one pos- 
sible hearth at Yaughan and only one possible interior hearth. There is no 
evidence for chimneys or wattle and daub or brick, and no evidence that any 
of the structures burned from open fires as would be expected with interior 
hearths. 



-Slaves did not cook or eat in a central kitchen nor was cooked food regular- 
ly brought into the quarters. No kitchen structure was found at any site, 
and particularly at completely excavated Site 38BK76. The dispersion of 
Colono cooking pots indicates that cooking and eating were conducted across 
the sites. 



-Slaves cooked and probably ate as individuals or in small groups. This is 
indicated by the small size of cooking pots, which would not have contained 
more than a quart at a time, and the single-serving size of the bowls. This 
is also supported by our conclusion that there was no central kitchen. 

-Slaves ate mostly stews, soups and gruel. Only one sherd of a shallow 
Colono bowl or "plate" was found at the sites, in contrast to the many Colono 
plates found in Charleston (Lise personal communication 1980). The remaining 
Colono vessels were bowls, cooking pots and a few examples of strainers and a 
lid handle possibly representing teapots. 

-Slaves ate less soup and stew at 38BK75. This is indicated by a shift in 
ceramic forms and types from site 38BK76 to 388K75, and by a slight increase 
in tableware at 38BK75 and 38BK245. 



341 



-Slaves usually ate with their fingers or with wooden spoons. There is very 
little evidence for metal forks, knives or spoons at the sites. Negative 
evidence indicates that the slaves must have used wood or other perishable 
materials to fashion eating utensils. Such materials do not make good forks 
or knives. 

What can these conclusions say about acculturation in the slave quarters? 
First, there appears to have been less change in foodways during the slave 
occupation than is shown by house types or many artifact types. The same 
foods, the same kind of cooking location, and probably the same hearth size 
or eating group size obtains from the earliest to latest periods. This is to 
be expected if foodways are less subject to change from outside. On the 
other hand, there is one change that may outweigh this seeming cultural 
inertia, and that is evidence for a decrease in the liquid or semi-liquid 
cuisine at 38BK245 and especially at 388K75. Chapter VIII discusses the 
changes in ceramic forms in some detail. 

Several statistically significant shifts in ceramic forms and types resulted 
in conclusions that Colono ceramics decreased and nonlocal ceramics increased 
over time. Within this trend significant changes in form also took place. 
Overall, cooking pots decreased during the later periods and flatware in- 
creased dramatically. The number of Colono cups and bowls increased relative 
to Colono cooking pots. Part of the decrease in cooking pots undoubtedly was 
made up for by iron kettles. The general trend indicates that the ceramics 
changed from those typical of liquid or semi -liquid foods to those more typi- 
cal of foods prepared to be eaten from flatwares. 

The shift from no permanent hearths at 38BK76 to a more permanent hearth with 
posts at 38BK75 has implications for cooking along with the shift from Colono 
pots to iron kettles. Posts would have allowed iron kettles (but not handle- 
less Colono pots) to be hung from a cross support. If the hypothesis con- 
cerning the use of a pot resting on a few stones is generally valid for 
388K76, then a permanent hearth with posts and iron kettles with legs and 
handles represents a change in how hearths were used. Unfortunately, this 
suggested change is only supported by two possible hearths without posts at 
38BK76, a lack of chimneys at all sites, and one permanent hearth with posts 
at 38BK75. It is also possible that some of the unidentifiable postholes at 
38BK76 may have represented the use of posts for temporary hearths at that 
site, although the two best candidates for hearths did not have posts. In 
any case, the impermanence of hearths at 38BK76 and the permanence of the 
hearth at 38BK75 are supported by the data even if the differential use of 
posts is not. 

A significant change from one set of forms to another and a less substanti- 
ated change from temporary hearths to more permanent hearths, when these 
changes imply new cooking and serving methods, is the kind of subtle change 
that reflects change in the non-material culture. The proper way of cooking 
and eating is one of the most traditional aspects of culture learned from 
childhood. Change in such a tradition is the best proof of acculturation. 



342 



XIII. CONCLUSIONS AND FUTURE RESEARCH QUESTIONS 

The historical research showed that Yaughan (38BK75 and 38BK76) and Curriboo 
(38BK245) plantations were owned by members of the Cordes family from 1737 to 
the mid-1830s. Although begun as secondary ventures ancillary to the fami- 
ly's holdings in the Parish of St. John, Berkeley, both plantations became 
residences as well as economic units in the mid to late eighteenth century. 
With their sale away from members of the family, they were either consoli- 
dated into larger holdings or subdivided. Studying the period from the 1740s 
to the 1820s, moreover, covered an entire cycle, from settlement through a 
frontier stage to residence and full occupation to sale and disintegration. 

Five hypotheses were set forth at the beginning of analysis to guide the re- 
search. These have been addressed throughout the report and are summarized 
below. 

Hypothesis 1 . The Colono ceramics recovered from sites 
3«BK/b, 38BK/6 and 38BK245 represent ceramics that were made 
by slaves who occupied the plantations, and that the slaves 
produced those wares for their own use. It is further hy- 
pothesized that the Yaughan and Curriboo Plantation samples 
are representative of the colonowares that were being made and 
used by African slaves in coastal South Carolina during that 
period. 

In Chapter VIII it was shown that two types of colonoware ceramics were pres- 
ent at Curriboo and Yaughan plantations. These were Colono and Catawba. It 
was concluded that Colono was made by and for slaves. Briefly, this conclu- 
sion was based on the large quantities of Colono in the slave quarters and 
its low percentage at the one extant owner's house; on the presence of vari- 
ous skill levels within the type (i.e. the Smoothed and Tooled Varieties); on 
the presence of unmarketable pots and objects; on the presence of unfired 
sherds at Curriboo; on the variety of objects recovered including pipes, 
strainers and marbles; and on Colono's resemblance to known slave made 
ceramics on the Caribbean islands. Further, the majority of Colono was not 
coil -made as native American ceramics were supposed to have been; the clays 
used to make Colono resembled locally available clay; and there are features 
at the sites which can only be satisfactorily explained as clay extraction 
pits. 

The only evidence which might have disproved the hypothesis was that no kilns 
were found. This, however, can be explained if the slaves fired Colono on 
the ground surface as is documented in West Africa. The presence of fine 
charcoal fragments throughout the stratigraphy of the site would support such 
an explanation. 

Comparisons with Colono and Catawba ceramics in the area and especially with 
the Charleston Museum collection clearly indicated that native Americans made 
Catawba, and that the method of manufacture and some of the forms of Colono 
resembled the Curriboo and Yaughan material. 



343 



Hypothesis 2 . Colonoware declined in importance at the 
plantations as time passed. Conversely, there was a trend 
toward greater dependence on non-1 ocally produced ceramics 
from the eighteenth to nineteenth centuries. 

Data presented in Chapter VIII showed a clear increase in the nonlocal 
ceramics and a decrease in Colono from 1740 to 1825. The relative amounts 
did not reach the Anglo-American percentages at the end of this period, but 
there was a clear trend in that direction. Reasons for this trend were 
hypothesized to have been an increase in the slave population, a resident 
owner at Yaughan, and intensification of agricultural activity which caused 
greater regimentation to be imposed on the slaves and allowed them less free 
time to pursue individual craft activities. Perhaps the most important rea- 
son for this trend was acculturation of the slaves themselves as illustrated 
by a shift in ceramic forms from jars and pots to more kettles, bowls and 
flatware. 



Hypothesis 3 . Patterns of artifact disposal on archaeological 
sites are culturally determined and can be discerned through 
careful analysis. Since African slaves were the products of 
radically different backgrounds than Anglo-Americans, the 
pattern of artifact disposal on African slave domestic sites 
should be discernible from that present on Anglo-American 
sites. The difference will be expressed in varying frequen- 
cies of the artifact categories that the sites share in com- 
mon, as well as the absence of certain categories. 

The patterns of artifact disposal were of two kinds, the relative amounts of 
artifact types and the location of trash disposal areas. Chapter IX has 
successfully shown that the relative amounts of artifact types differed 
significantly between the slave sites studied here and the Anglo-American 
sites investigated by South (1977a). This was especially true for the fre- 
quencies of Kitchen and Architecture group artifacts. Slave sites were 
clearly more heavily weighted towards Kitchen Groups artifacts and less 
towards Architecture Group artifacts. 

The location of disposal areas also appeared to differ from South' s (1977a) 
"odorimetric" scale which states that organic refuse will be deposited away 
from residential structures. Trash features at all sites showed that this 
was not the case in the slave quarters. 

Hypothesis 4 . The earliest construction technique in use 
within tFie Yaughan Plantation slave quarters involved wall 
trench construction coupled with the use of a few individual 
postholes. As time passed, that mode was superceded by the 
use of entirely individually set posts. This construction 
sequence represents a transition from construction techniques 
used in the Caribbean and/or Africa and reflects the greater 
acculturation of slaves into the Anglo-American sub-culture as 
time passed. 



344 



The archaeological and historical evidence both pointed to the fact that 
trench construction was superceded by posthole construction. Further, and 
perhaps more importantly, there appeared to be a shift from mudwall construc- 
tion to frame construction which correlated closely with the shift in found- 
ation type. The form of the structures also shifted from rectangular to more 
square or to a more Anglo-American pattern. Although detailed data could not 
be developed on African and Caribbean structures, the literature did seem to 
support the hypothesis. 

Hypothesis 5 . The earliest construction technique within 
Curriboo Plantation involved wall trench construction coupled 
with the use of a few individual postholes. As time passed, 
that mode was superceded by the use of brick pier construction 
for major plantation outbuildings. This construction sequence 
represents a transition from a frontier pattern to a more per- 
manent settlement pattern and a successful adaption to pre- 
vailing economic trends. 

It was shown at Curriboo Plantation that the early structures in the slave 
quarters and at the location of the hypothesized office structure were of 
trench construction. The brick piered office was also superimposed on the 
remains of a trench foundation, naval stores processing structure. Unfor- 
tunately, no other restricted-use structures were found for the later periods 
and this hypothesis could not be adequately tested. 

The slave occupation at Yaughan falls into three periods, from the 1740s to 
about 1780 characterized by isolation from Anglo-American society, from 1780 
to about 1805 characterized by a dramatic increase in population and in- 
creased contact with the dominant Anglo-American culture, and from 1805 to 
the 1820s characterized by ever increasing contact with the outside world and 
a decline in the economic condition and population of the plantation. Occu- 
pied from the 1740s to the turn of the eighteenth century, Curriboo was 
characterized by conditions similar to those at Yaughan from 1780 to 1805. 

Although the demographic and cultural parameters described from historical 
research do not demonstrate the existence of an autonomous black culture in 
eighteenth century plantation life, they argue powerfully that the social 
environment was ripe for such a culture to flourish. Direct testimony from 
eighteenth century slaves is lacking in the historical record, and all too 
frequently, historians have read the evidence from nineteenth century slaves 
back into the Colonial Period. The historical documentation has shown very 
clearly that the eighteenth century plantation was different from the nine- 
teenth century plantation. Life in the slave quarter was one of these dif- 
ferences, and the archaeology that has ensued from this project represents 
direct statements from long dead Africans and demonstrates again the dif- 
ference between the colonial and antebellum plantation systems. 

Archaeological investigations at Yaughan and Curriboo indicated that much of 
value to the study of slave life could be gathered from archaeological data. 
With the conclusions on slave life at the plantations established by the his- 
torical documents, certain facets of slave life could be studied as internal 



345 



culture change and acculturation within a limited area. Archaeology support- 
ed the essential outline of the three periods of acculturation and offered 
data and conclusions not available from the historical sources alone. Among 
the cultural patterns established from the material remains of the slave 
quarters were the attribution of Colono ceramics to black slaves and Catawba 
ceramics to Indian trade wares. Artifact types indicated increasing accept- 
ance of new types by the slaves at Yaughan and Curriboo. Overall artifact 
patterns and mean ceramic dates also supported major shifts in the acceptance 
of whole different functional categories of artifacts. Architecture shed 
light on how the slaves lived and illustrated acculturation on a deeper and 
more pervasive level of culture. Studies of slave foodways showed a change 
from being less Anglo-American to more Anglo-American in food preparation and 
consumption patterns. 

In conclusion, it has been possible to illustrate acculturation in the ar- 
chaeological record through the study of artifacts, minor artifact types, 
artifact patterns, architecture and foodways. Many questions have been 
raised concerning lifeways in the slave quarter and the acculturation of an 
important segment of the American population. Much more work needs to be 
done in the archaeological study of slavery. Possible areas of concern are: 

-What differences in the character of plantations from the eighteenth and 
nineteenth centuries caused the shifts noted in the archaeology of these two 
periods? Facile overgeneralizations that rice plantations used the task 
system of labor and cotton plantations used the gang system do not get to the 
heart of the matter. Thorough, site specific history is necessary to control 
not only the when and who of history, but also the deeper meanings contained 
in historical documents. 

-How can changes documented in the historic records be identified archaeologi- 
cally? This is a problem for all of historical archaeology and involves 
examination of what kinds of things are preserved archaeological ly that may 
reflect historical records and how these things should be analyzed to study 
or examine the non-material aspects of culture. 

-Does the Carolina SI ave Arti f act Pattern hold up throughout South Carolina 
during the eighteenth century? Such a pattern might be more properly named 
the Berkeley County Slave Artifact Pattern if other slave sites in South 
Carolina are examined more closely. How does the pattern compare with slave 
sites in other regions and how does it change in the nineteenth century? 
Without such a pattern for comparison, differences in the archaeological 
record cannot be distinguished. 

-Are there any patterns which remain the same throughout Afro-American his- 
tory and that are identifiable in the archaeological record? Studies in 
freed slave and twentieth century black neighborhoods are presently being 
conducted. The trend of artifact patterns as presently set up tends to 
indicate that such patterns on more recent sites would be indistinguishable 
from Anglo-American patterns, however, there may be patterns not yet devel- 
oped which would show ethnic differences in the nineteenth and twentieth 
centuries. 



346 



-Is there a difference between acculturation in rural and urban slave set- 
tings that can be identified archaeological ly? Colonoware forms have already 
been suggested as a possible avenue for such research. 

-What was the cultural heritage brought from Africa? This project neces- 
sarily starts with slaves already acculturated to some extent, and the full 
range of acculturation could not be examined because of the almost total lack 
of comparative archaeological data on African groups from 1500 to 1850. 

-Can social groups be characterized as dependent or having self-esteem, and 
if so, can this be determined archaeologically? Studying groups of people 
from slavery to freedom could provide such insights and provide time depth to 
psychological anthropology. 

We feel it is appropriate that a project such as this ends with perhaps more, 
and more interesting, questions to be answered than it began with. While 
some questions have been answered and many more have been left unanswered it 
is hoped that the data presented here will be used by other researchers to 
examine more complex anthropological themes. 



347 



REFERENCES CITED 

Adams, William H., and Linda P. Gaw 

1977 A model for discerning time lag of ceramic artifacts. Northwest 
Anthropological Research Notes 11(2) :218-231 . 

Anderson, David G. 

1977 A history of prehistoric archeological investigations in the 
Coastal Plain of South Carolina. South Carolina Antiquities , Vol. 
9, No. 2. 

Anderson, David G., Sammy T. Lee, and A. Robert Parler, Jr. 

1979 Cal Smoak, archaeological investigations along the Edisto River in 
the coastal plain of South Carolina. Occasional Papers of the 
Archeological Society of South Carolina, Inc. T. 

Anthony, Ronald W.. 

1979 Descriptive analysis and replication of historic earthenware: 

colono wares from the Spiers Landing site, Berkeley County, South 
Carolina. The Conference on Historic Site Papers 1978 13:253-268. 

Archive News 

1968 South Carolina Historical Magazine 69. 

Ascher, Robert and Charles H. Fairbanks 

1971 Excavation of a slave cabin: Georgia, U.S.A. Historical 
Archaeology 5:3-17. 

Baker, Steven G. 

1972 Colono-Indian Pottery from Cambridge, South Carolina with comments 
on the historic Catawba pottery trade. IAA Notebook Vol . IV, no.- 

1:3. 

Baker, Vernon G. 

1978 Black Lucy's Garden. Papers of the Robert S. Peabody Foundation 
8. Andover. 

Ball , Joseph 

1973 Ceramic sequence at Becan, Campeche, Mexico . University Micro- 
films, Ann Arbor. Unpublished Ph.D. dissertation, Tulane 
University. 

1977 Book reviews. American Antiquity 42:661-663. 

1979 On data, methods, results, and reviews: a reply to Michael Smith. 
American Antiquity 44:828-831. 

Bartovics, Al 

1978 Personal correspondence with Christine Johnson. 



348 



Bentley, William George 

1977 Wealth distribution in colonial South Carolina. Ph.D. disser- 
tation, Georgia State University, Atlanta. 

Blalock, Hubert M. 

1960 Social Statistics. McGraw Hill, New York. 

Bonner, James C. 

1945 Plantation architecture of the lower south on the eve of the Civil 
War. Journal of Southern History 11:370-388. 

Boucher, Edward Joseph 

1948 Vestrymen and churchwardens in South Carolina, 1706-1778. 

Unpublished M.A. thesis, Department of History, University of 
South Carolina. 

Bower, Beth Anne and Byron Rushing 

1980 The African meeting house: the center for the 19th century 

Afro-american community in Boston. In archaeol ogi cal I perspecti ves 
on ethnicity in America , edited by R. Schuyler, pp. 69-75, Baywood 
Publishing Co., Inc., Farmingdale, N.Y. 

Braley, Chad 0. 

1980 The mean ceramic date formula and antebellum slave sites: chrono- 
logy versus culture processT Paper presented at the 21st Annual 
Meeting of the Conference on Historic Site Archaeology, Wil- 
mington, N.C. 

Brasington, Dal ton K. 

1977 A historical geography of South Carolina's inner coastal plain. 
M.A. thesis, University of South Carolina, Columbia. 

Brewster, Lawrence Fay 

1947 Summer migrations and resorts of South Carolina's Low-country 
planters 7 Duke University Press, Durham, North Carolina. 

Bridges, Sarah T. and Bert Sal wen 

1980 Weeksville: the archaeology of a black urban community. In 

archaeological perspectives on ethnicity in America , edited" by R. 
Schuyler, pp. 38-47, Baywood Publishing Co. Inc., Farmi ngdal e , 
N.Y. 

Brockington, Paul Jr. 

1980 Cooper River Rediversion Canal archeological survey. Institute of 
Archeology and Anthropology, University of South Carolina, 
Research Manuscript Series No. 169T 

Brown, Philip M. 

1975 Early Indian trade in the development of South Carolina: poli- 
tics, economics, and social mobility during the proprietary peri- 
od, 1670-1719. South Carolina Historical Magazine 76:118-128. 



349 



Brunei", Edward M. 

1956 Primary group experience and the process of acculturation. 
American Anthropologist 58:605-623. 

Bull en, Adelaide K. and Ripley P. Bull en 

1945 Black Lucy's garden. Bulletin of Massachusetts Archaeological 
Society 6:17-28. 

Bull en, Ripley P. and H. Bruce Green 

1970 Strati graphic tests of Stal lings Island, Georgia. Florida 
Anthropologist 23(l):8-28. 

Bureau of the Census, Department of Commerce 

1943 U.S. census of population: 1940 (Vol. II). Characteristics of 
the population (Part 6), Pennsylvania-Texas . Government Printing 
Office, Washington, D.C. 

1952 U.S. census of population: 1950 (Vol. II). Characteristics of 
the population (Part 40), South Carolina . Government Printing 
Office, Washington, D.C. 

1963 U.S. census of population: 1960 (Vol . I). Characteri sti cs of the 
population (Part 42), South Carolina . Government Printing Office, 
Washington, D.C. 

1973 1970 census of population (Vol. I). Characteristics of the popu- 
lation (Part 42), South Carolina . Government Printing Office, 
Washington, D.C. 

Bureau of the Census, Department of Commerce and Labor 

1908 Heads of Families at the First Census of the United States taken 
in the Year 1790 . Washington, D.C. 

Bureau of the Census, Department of Commerce 

1895 Report on the statistics of agriculture in the United States at 
the eleventh census: 1890T Government Printing Office, Washing- 
ton, D.C. 

1901 U.S. census of population: 1900 (Vol . I). Population . Govern- 
ment Printing Office, Washington, D.C. 

1902 U.S. census: 1900 (Vol. VI). Agriculture (Part 2), crops and 
i rri gati orT Government Printing Office, Washington, D.C. 

1913 U.S. census of population: 1910 (Vol. III). Population, reports 
by states, Nebraska-WyomingT government Printing Office, Washing- 
ton, D.C. 

1913 U.S. census of population: 1910 (Vol. VII). Agriculture, reports 
by states, Nebraska-Wyoming^ Government Printing Office, 
Washington, D.C. 

1922 U.S. census of population: 1920 (Vol. III). Population, 

composition and characteri sti cs~"of the population by states . 
Government Printing Office, Washington, D.C. 



350 



1922 U.S. census of population: 1920 (Vol. VI). Agriculture, reports 
by states (Part 2), southern states . Government Printing Office, 
Washington, D.C. 

1932 U.S. census of population: 1930 (Vol. II). Agriculture (Part 2), 
The southern states! Government Printing Office, Washington, D.C. 

1932 U.S. census of population: 1930 (Vol. III). Population, reports 
by states (Part 2), Montana-Wyoming . Government Printing Office, 
Washington , D.C. ~ 

1942 U.S. census of population: 1940 (Vol. I). Agriculture, state 
reports (Part 3), statistics for counties . Government Printing 
Office, Washington, D.C. 

C. W. 

1755 The indigo plant discussed. The Gentleman's Magazine , Sylvanus 
Urban, ed., pp. 201-203. (Carol iniana Library, Columbia, South 
Carolina). 

Caldwell , Joseph R. 

1965 Primary forest efficiency. Southeastern Archaeological 
Conference , Bulletin 3 . 

Carrillo, Richard F. 

1980 Green Grove Plantation: Archaeological and historical research at 
the Kinlock Site (38CH109). Ms. on file, South Carolina 
Department of Highways and Public Transportation, Columbia, South 
Carolina. 

Charleston Probate Records 

1671- Record of Wills. South Carolina Department of Archives and 
1868 History, Columbia; Charleston County Public Library, Charleston; 
Charleston Probate Office, Charleston. 

Cordes, Rebecca 

1839- Vol. 43, pp. 680-683. 

1845 

Cordes, Alexander Watson 

1974 The Cordes' on that side . . . Transactions of the Huguenot 
Society of South Carolina 79:86. 

Cordes, Anthony 

1671- Vol. 1, pp. 37-38. 

1724 

Cordes, Charlotte 

1826- Vol. 37, pp. 238-241. 

1834 



351 



Cordes, Henrietta Catharine 

1760- Vol. 10, pp. 442-444. 
1767 

Cordes, Isaac 

1740- Vol. 5, pp. 406-409. 

1747 

Cordes, James Paul 

1776- Vol. 18, pp. 203-205. 

1784 

Cordes, James 

1786- Vol. 23, Book B, p. 414. 

1793 

Cordes, John 

1752- Vol. 7, pp. 582-584. 

1756 

Cordes, John 

1793- Vol. 27, Book C, pp. 770-772. 

1800 

Cordes, Samuel 

1793- Vol. 26, Book B, pp. 504-507. 

1800 

Cordes, Samuel 

1856- Vol. 48, p. 162. 

1862 

Cordes, Thomas 

1747- Vol. 6, pp. 141-145. 

1752 

Cordes, Thomas 

1760- Vol. 10, Book B, pp. 450-452. 

1767 

Cordes, Thomas Jr. 

1793- Vol. 27, Book C, pp. 960-961. 

1800 

Lequeux, Peter 

1771- Vol. 14, pp. 107-110. 

1774 

Lucas, Jonathan 

1834- Vol. 40, Book A, pp. 461-470. 

1839 



352 



Will of Solomon Clarke 

1845- Vol.45, Book A, pp. 771-773. 

1851 

Charleston Probate Records 

1736- Inventories. South Carolina Department of Archives and History, 
1776 Columbia; Charleston County Public Library, Charleston; Charleston 
Probate Office, Charleston. 

Cord[e]s, John 

1756- 22 January 1757, Inventories, S, pp. 22-28. 

1758 

Cordes, Col. Thomas 

1748- Vol. B, pp. 124-129. 

1751 

Cordes, Henrietta Catharine 
10 Jan Vol. W, pp. 221-222. 

1765 

Cordes, Isaac 

1732- 9 August 1745, Vol. 67A, pp. 316-332. 

1746 

Cordes, Thomas 

n.d. Vol. V, pp. 492-494. 

Clonts, John B. 

1979 Some long overdue thoughts on faunal analysis. Ms., Western 
Archaeological Center, National Park Service. 

Clowse, Converse D. 

1971 Economic beginnings in colonial South Carolina, 1670-1730 . South 
Carolina Tricentennial Commission, University of South Carol ina 
Press, Columbia. 

Comptroller General Records 

1778- Revolutionary Bonds. South Carolina Department of Archives and 
1781 History, Columbia. 

1824 Tax Returns. South Carolina Department of Archives and History, 
Columbia. 

Cordes, Ann 
5 May 1778 

Cordes, James 
1781 

Cordes, John 
5 Apr 1780 



353 



Cordes, Samuel 
1782 

Cooper, Joseph 

1979 Personal communication, Landowner, Berkeley County, S.C. 

Cordes, John, Estate of 

1757- Account Book. W. P. A. Transcript of original in possession of 
1798 Philip G. Porcher, 1932-1937. Library of the College of Charles- 
ton, Charleston. 

Cordes 

1974 The Cordes' on the other side. South Carolina Historical Magazine 
75. 

Counts, Charles 

1981 Personal communication. Potter in Rising Fawn, Georgia. 

Court of Common Pleas 

1791- Judgement Rolls. South Carolina Department of Archives and 
1910 History, Columbia. 

Cowgill , George 

1977 The trouble with significance tests and what we can do about it. 
American Antiquity 42, 3:350-368. 

Crokatt, James 

1746 Observations concerning indigo and cochineal, also on silk, rice, 
pitch, tar, and turpentine" ! London. 

1747 Further observations intended for improving the culture and curing 
of indigo in South Carol inaT London . " 

Curtin, Philip D. 

1969 The Atlantic slave trade; a census . University of Wisconsin 
Press, Madison and London. 

de Beauvais, Raseau 

1770 L'art de L'indigotier . France. 

DePratter, Chester B. 

1975 The Archaic in Georgia. Early Georgia 3(1): 1-16. 

1977 Environmental changes on the Georgia coast during the prehistoric 
period. Early Georgia 5(1 and 2). 

Deetz, James 

1977 In small things forgotten . Anchor Press, Garden City, New York. 

Douglas, R. W. and Susan Frank 

1972 A history of glassmaking . G. T. Foults and Co., (no place of 
publication). 



354 



Downie, N. M. and R. W. Heath 

1974 Basic statistical me thods , fourth edition. Harper and Row, New 
YSrT. 

Drewel , Henry 

1980 Personal communication. Art historian, University of Indiana. 

Drucker, Lesley M. 

1979 Spiers Landing: a socioeconomic study of an undocumented late 

eighteenth century site in Berkeley County, South Carolina . Paper 
presented at the Society for Historical Archaeology Conference , 
Nashville. 

Drucker, Lesley M. and Ronald W. Anthony 

1979 The Spiers Landing Site: archaeological investigations in Berke- 
ley County, South Carolina^ Carolina Archaeological Services, 
Columbia, South Carolina. 

Dubose, Samuel 

n.d. Reminiscences of St. Stephen's Parish, Craven County and notices 
of her old homesteads. In A contribution to the history of the 
Huguenots of South Carolina , compiled by T. Gail lard Thomas, M.D., 
pp. 35-86. R. L. Bryan, Columbia. Originally published 1852, 
reprint n.d. 

Eaton, Clement 

1975 A history of the old South; the emergence of a reluctant nation . 
Macmil lan, New York. 

Ebanks, R. C. 

1974 Letter to the editor. Jamaica Journal 8(2 & 3): 2. 

Eg! off, Keith 

1980 Colonial plantation hoes of Tidewater Virginia. Research Report 
Series No . 1 . 

El kins, Stanley 

1960 Slavery; a problem in American institutional and intellectual 
1 i f e . University of Chicago Press, Chicago. 

Fagg, William and John Pic ton 

1978 The potter's art in Africa . British Museum Publications Ltd., 
London. 

Fairbanks, Charles 

1962 A Colono-Indian ware milk pitcher. Florida Anthropologist XY(4). 

1974 The Kingsley slave cabins in Duval 1 County, Florida, 1968. The 
Conference on Historic Site Archaeology Papers 1972 7:62-93. 

Fairhurst, Janet 

1974 Sainte Genevieve. Living History, August: 56-59. 



355 



Ferguson, Lei and 

1971 South Appalachian Mississippian . Ph. D. dissertation, University 
of Nortn Carolina. University Microfilms, Ann Arbor. 

1977a Looking for the "Afro" in Colono-Indian pottery. Conference on 
Historic Site Archaeology Papers 12:68-83. Columbia. 

1977b An archaeological -historical analysis of Fort Watson. In Research 
strategies in historical archeology , edited by S. South, pp. 
41- /Z, Academic Press, New York 

1980 The "Afro-" in Colono-Indian pottery. In Archaeological 

perspectives on ethnicity in America , edited by R. Schuyler, pp. 
14-28, Baywood Publishing Company, Inc., Farmingdale, N.Y. 

Fernald, Merritt L. and Alfred C. Kinsey 

1958 Edible wild plants of eastern north America . Second edition, 
revised by R. C. Rollins. Harper and Row, New York. 

Fish, Paul 

1978 Consistency in archaeological measurement and classification: a 
pilot study. American Antiquity 44: 828-831. 

Fogel and Engerman 

1974 Time on the cross: the economics of American Negro slavery . 
Little Brown and Company, Boston. 

Foust, James D. 

1975 The yeoman farmer and westward expansion of U.S. cotton 
production" ! Arno Press, New York. 

Freed, Stanley A. 

1957 Suggested type societies in acculturation studies. American 
Anthropologist 59:55-68. 

Friedlander, Amy 

1975 Indian slavery in proprietary South Carolina . M.A. thesis, Emory 
University, Atlanta. 

1979 Carolina Huguenots; a study in cultural pluralism in the low 
country, 1679-1 761T Ph.D. dissertation, Emory University, 
Atlanta. 

1980 Historic research synthesis: Yaughan and Curriboo Plantations. 
Ms. on file, Soil Systems, Marietta, Georgia. 

Gail lard, Phillip 

1979 Personal communication. Landowner, Berkeley County, S.C. 



356 



Garrow, 
1979 



1981 



Patrick H. 

Proposal to perform testing and data recovery at five historic 
sites, Cooper River Rediversion Canal, Berkeley County, South 
Carolina. Ms. on file, Soil Systems, Inc., Marietta, Georgia. 



Washington D.C. Civic Center mitigation program. 
Soil Systems, Inc., Marietta, Georgia. 



Ms. on file, 



Garrow, Patrick H. and Thomas R. Wheaton, Jr. 

1979 Documented recommendations for additional investigations of his- 
toric sites to be impacted by construction of the Cooper River 
Rediversion Canal, Berkeley County, South Carolina. Ms. on file, 
Soil Systems, Inc., Marietta, Georgia. 

Garrow, Patrick H. and G. Ishmael Williams 

1980 An initial investigation of the archaeological and historic re- 
sources which would be affected by the continued operation and 
maintenance of the Atlantic Intracoastal Waterway, Little River to 
Port Royal Sound, South Carolina. Ms. on file, Soil Systems, 
Inc., Marietta, Georgia. 

Gartley, Richard T. 

1979 Afro-Cruzan pottery - a new style of colonial earthenware from St. 
Croix. Journal of the Virgin Islands Archaeological Society 
8:47-61. 

Geismar, Joan H. 

1980 Skunk Hollow: a preliminary statement on archaeological 
investigations at a 19th century black community. In 
Archaeological perspectives on ethnicity in America , edited by R. 
Schuyler, pp. 49-60, Baywood Publishing Company Inc., Farmingdale, 
N.Y. 

Genovese, Eugene D. 

1974 Roll, Jordan, roll; the world the slaves made. Vintage Books, New 
York-; 

Glassie, Henry 

1968 Pattern in the material folk culture of the eastern United States . 
University of Pennsylvania Press, Philadelphia. 

1975 Folk housing in middle Virginia . University of Tennessee Press, 
Knoxville. 

Good, Mary Elizabeth 

1972 Guebert Site: An 18th century, historic Kaskaskia Indian village 
in Randolph County, Illinois. Central States Archaeological So- 
cieties Memoir II. 



357 



1976 Glass Beads from the First Hermitage. In An Archaeological and 
Historical Assessment of the First Hermitage, edited by Samuel D. 
Smi th . Tennessee Department of Conservation, Division of Ar- 
chaeology , Research Series 2:237-248. 

Gray, Lewis Cecil 

1933 History of agriculture in the southern United States to 1860 . 
Carnegie Institution of Washington, Washington, D.C. 

Gray, Marlesa 

1979 Personal communication. Staff archaeologist, WAPORA Corporation. 

Grayson, Donald K. 

1973 On the methodology of faunal analysis. American Antiguity 
38(4): 432-438. 

Greene, Evarts B.and Virginia D. Harrington 

1966 American population before the Federal Census of 1790 . Peter 
Smith, Gloucester, Mass. 

Griff en, James B. 

1952 Radiocarbon dates for the Eastern United States, in Archeology of 
the Eastern United States . University of Chicago, Chicago. 

1967 Eastern north American archaeology: A summary. Science 
156:175-91. 

Griffiths, Dorothy 

1978 Use marks on historic ceramics: a preliminary study. Historical 
Archaeology 12:68-81. 

Guidoni, Enrico 

1978 Primitive architecture . Harry N. Abrams, New York. 

Gutman, Herbert G. 

1976 The black family in slavery and freedom, 1750-1925 . Pantheon 
Books, New York. 

Haberland, W. 

1963 On the classification and systematics of precolumbian ceramics in 
America. Ceramica de Cultura Maya , l:no. 2-3:8-22. 

Hally, David J. 

1975 The Mississippian Period. Early Georgia 3(l):37-52. 

Hamilton and Marrinan 
1979 

Handler, Jerome 

1963 Pottery making in rural Barbados. Southwestern Journal of 
Anthropology 19:314-334. 



358 



1964 Notes on pottery making in Antiqua. Man, a Record of Anthropo- 
logical Science . 19:314-334. 

Handler, Jerome and Frederick W. Lange 

1978 Plantation slavery in Barbados: an archaeological and historical 
i nvesti gati on . Harvard University Press, Cambridge. 

Hannum, Eleanor Clarke 

1970 The parish in South Carolina, 1706-1868 . Unpublished M.A. thesis, 
Department of History, University of South Carolina. 

Harris, R. King and Inus M. Harris 

1967 Trade Beads, Projectile Points, and Knives. In A pilot study of 
Wichita Indian archaeology and ethnohi story , edited by Robert E. 
Bell and others, pp. lZ9-lb8. Manuscript report, Southern Me- 
thodist University. 

Haskell, Helen 

1981 The Middleton Place privy house. Institute of Archeology and 
Anthropology Popular Series 1 , Columbia, S.C. 

Haywood, C. Robert 

1959 Mercantilism and South Carolina agriculture, 1700-1763. South 
Carolina Historical Magazine 60:15-27. 

Henry, Susan 

1980 Physical, spatial, and temporal dimensions of Colono Ware in the 
Chesapeake 1600-18001 Unpublished M.A. thesis, Catholic 
University of America. Wash. O.C. 

Herndon, G. Melvin 

1979 Forest products of colonial Georgia. Journal of Forest History 
July: 130-135. 

Herold, Elaine B. and Stanley G. Knick, III 

1978 An archeological survey of the proposed Santee-Cooper Power Plant 
site, St. Johns Parrish, Berkeley County. Ms. on file, Charleston 
Museum, Charleston. 

Higgins, W. Robert 

1964 Charles Town merchants and factors dealing in the external Negro 
trade, 1735-1775. South Carolina Historical Magazine 65:205-217. 

Hill , James N. 

1970 Broken K. Pueblo: prehistoric social organization in the American 
Southwest. University of Arizona, Anthropological Papers , 18. 

HUliard, Sam B. 

1969 Hogmeat and corn pone: food habits in the antebellum south 
American Philosophical Society Procedures 113( 1 ) : 1-13. 



359 



Hirsch, Arthur Henry 

1928 The Huguenots of colonial South Carolina . Duke University Press, 
Durham, North Carolina. 

Honigman, John J. 

1973 Handbook of social and cultural anthropology . Rand McNally and 
Company, Chicago. 

Hudson, Charles M., Jr. 

1965 The Catawba nation: a social history . Unpublished Ms. Ph.d. 
dissertation. University of North Carolina, Chapel Hill. 

Jelks, Edward B. 

1973 Archaeological explorations at Signal Hill, Newfoundland, 
1965-1966. Occasional Papers in Archaeology and History 7 , 
Ottawa. 

Johnson, Willis G. and G. 0. Brown, eds. 

1903 The poultry book . Doubleday, New York. 

Jones, Olive 

1971 Glass bottles: pushups and pontil marks. Historical Archaeology 
5:62-73 

Jones, Alice Hanson 

1980 Wealth of a nation to be . Columbia University Press, New York. 

Juricek, John T. 

1964 The Westo Indians. Ethnohi story 11:134-73. 

Keesing, Roger M. and Felix M. Keesing 

1971 New perspectives in cultural anthropology. Holt, Rinehart and 
Winston Inc., New York. 

Kelso, William 

1976 The colonial silent majority: tenant, servant and slave settlement 
sites at Kingsmill, Virginia . Paper presented at the /bth annual 
meeting American Anthropological Association, Wash. D.C. 

1979 Captain Jones's Wormslow. 

1980 Map of Littletown Quarter Investigations, personal 
correstpondence. 

Kidd, Kenneth E. 

1979 Glass bead-Making from the Middle Ages to the early 19th century. 
History and Archaeology 30 , Parks Canada, Ottawa. 

Kniffen, Fred and Henry Glassie 

1966 Building in wood in the eastern United States: a time-place 
perspective. Geographical Review 56:48-65. 



360 



Kniffen, Fred 

1965 Folk Housing: key to diffusion. Annals of the Association of 
American Geographers 55:549-576. 

Lathrap et al 

1956 An archaeological classification of culture contact situations. 
American Antiquity XXII, no. 2, part 2:1-30. 

Lawson, John 

1967 A new voyage to Carolina . Edited by H. T. Lefler. University of 
North Carolina, Chapel Hi 1 1 . 

Leavitt, Genevieve 

1980 Personal correspondence. Graduate Student, College of William and 
Mary. 

Lees, William 

1979 Pattern and meaning of Colono-Indian ceramics at Limerick 

plantation, South Carolina. Pattern and Meaning 1(4), Columbia, 
South Carolina. 

Lees, William and Kathryn M. Kimery-Lees 

1978 The function of Colono-Indian ceramics: insights from Limerick 
plantation. Historical Archaeology 12: 

Lewis, Kenneth 

1973 An archaeological consideration of the frontier. Papers in An- 
thropology 14 : (no. 1):84-103. 

1977 Sampling the archeological frontier: regional models and component 
analysis. In Research strategies in historical archeology , edited 
by Stanley South, pp. 151-202. Academic Press, N.Y. 

1979 Excavating a colonial rice plantation. Early Man , Autumn, 
1979:13-17 : 

1981 Abandonment vs. discard behavior on a postbell urn plantation: The 
Mi ddleton Place privy . Paper presented at the 22nd annual Histor- 
ic Sites Conference, Columbia, South Carolina. 

Lewis, Kenneth and Helen Haskell 

1980 Hampton II: Further archeological investigations at a Santee Ri- 
ver rice plantation. Research Manuscript Series, No. 161 : Colum- 
bia, South Carolina. 

Lof strom, Edward Urling 

1976 A seriation of historic ceramics in the midwest, 1780-1870 . Paper 
presented at the Joint Plains Midwest Anthropological Conference . 

Long, Bobby M. 

1980 Soil Survey, Berkeley County, South Carolina . U.S. Soil 
Conservation Service. 



361 



Longacre, William A. 

1970 Archaeology as anthropology: a case study. University of Arizona, 
Anthropological Papers , 17. 

Lucas Family 

1792- Papers. South Carolina Historical Society, Charleston. 
1906 

MacNeish, Richard, Fredrick Peterson, and Kent Flannery 

1970 The prehistory of the Tehuacan Valley , vol . 3, Ceramics. Uni- 
versity of Texas Press, Austin. 

Main, Jackson Turner 

1965 The social structure of revolutionary America . Princeton Uni- 
versity Press, Princeton. 

Matthewson, R. Duncan 

1973 Archaeological analysis of material culture as a reflection of 
sub-cultural differentiation in 18th century Jamaica. Jamaica 
Journal 7 (1 & 2):25-29, 6:54-56. 

Martin, Alexander C. 

1972 Weeds . Golden Press, New York. 

Matheny, R. 

1970 The ceramics of Aguacatal, Campeche, Mexico . New World Ar- 
chaeological Foundation, Provo. 

Meggers, Betty J. 

1975 The transpacific origin of Mesoamerican civilization: a 
preliminary review of the evidence and its theoretical 
implications. American Anthropologist 77:1-27. 

Michie, James L. 

1967 Fluted points of the South Carolina coast. The Chesopiean 
5(2):54. 

Miller, George L. 

1980 Classification and economic scaling of 19th century ceramics. 
Historical Archaeology 14:1-41. 

Miller, J. Jefferson and Lyle M. Stone 

1970 Eighteenth-century ceramics from Fort Michil imackinac. Smithsoni- 
an Institution Press, Smithsonian Studies in History and Technolo- 
gy No. 4 . Washington D.C. 

Milling, Chapman J. 

1969 Red Carolinians . University of North Carolina, Raleigh. 

Mills, Robert 

1826 Statistics of South Carolina. Hurl but and Lloyd, Charleston. 



362 



Misenhelter, Jane Searles 

1977 St. Stephen's Episcopal Church, St. Stephen, S. C. By the author, 
Bonneau, South Carolina. 

Moore, Jamie W. 

1979 The Lowcountry in economic transition: Charleston since 1865. 
South Carolina Historical Magazine 80:156-171. 

Morgan, Philip D. 

1977 The development of slave culture in eighteenth century plantation 
America . Unpublished Ph.D. dissertation, University College, 
London. 

Mull ins -Moore, Susan 

1979 The southern coastal plantation: view from St. Simons Island, 
GeorgT a . Paper presented at the 12th annual meeting, Society for 
Historical Archaeology, Nashville. 

1980 A preliminary discussion of antebellum plantations on the Georgia 
coast . Paper presented before the Society for Historical Archaeo- 
logy Meeting, Albuquerque, New Mexico. 

Munsell Soil Color Charts 

1975 Kollmorgen Corp., Baltimore. 

Noel Hume, Ivor 

1962 An Indian ware of the colonial period. Quarterly Bulletin of the 
Archaeological Society of Virginia 17(1): 2-4. 

1972 Creamware to Pearl ware: A Williamsburg perspective. In Ceramics 
in America . Edited by Ian Quimby, University Press of Virginia, 
Charlottesville. 

1978 A guide to artifacts of colonial America. Alfred A. Knopf, New 
TorT: 

01 sen, Stanley J. 

1963 Dating early plain buttons by their form. American Antiquity 
28:no. 4:551-554. 

Orvin, Maxwell Clayton 

1973 Historic Berkeley County, South Carolina, 1671-1900 . By the 
author, Charleston. 

Otto, John Solomon 

1975 Status differences and the archaeological record - a comparison of 

p ]^ nter ' overseer ' and s l ave s1tes from Cannon's Point Plantation 
(1794-1861), St. Simons Island, Georgia . Ph.D. dissertation, 
University of Florida. University Microfilms, Ann Arbor. 

1977 Artifacts and status differences — a comparison of ceramics from 
planter, overseer, and slave sites on an antebellum plantation. 
In Research Strategies in Historical Archeology , edited by Stanley 
South, pp. 91-118. Academic Press, New York. 



363 



1980 Race and class on antebellum plantations. In Archaeological 

perspectives on ethnicity in America , edited by R. S. Schuyler, 
pp. 3-13. Baywood Publishing Company, Farmingdale, N.Y. 

Outlaw, Abbitt, Merry, Beverly Bogley and Alain C. Outlaw 

1977 Rich man, poor man: status differentiation in two 17th century 
ceramic assemblages from KingsmilH Paper presented at 10th 
annual meeting, Society for Historical Archaeology, Ottawa. 

Palmer, Arlene M. 

1976 A Winterthur guide to Chinese expor t po rcelain . Crown Publishers, 
Inc., N.Y. 

Peebles, Christopher S. 

1979 Of types and varieties in book reviews: a rejoinder to Michael E. 
Smith. American Antiquity 44:826-827. 

Peterson, Eugene T. 

1963 Clay pipes: a footnote to Mackinac's history. Mackinac History, 
Leaflet Number 1 , Mackinack Island State Park Commission, Mackinac 
Island, Michigan. 

Petty, Julian J. 

1975 The growth and distribution of population in South Carolina . 
State Council for Defense, Industrial Development Committee, 
Columbia, South Carolina. Reprint of 1943 edition. 

Peyre, Thomas Walter 

1834- Journal. South Carolina Historical Society, Charleston. 
1850 

Pineville Police Association 

1823- Records. South Carolina Historical Society, Charleston. 
1843 

Piatt, J. Lamar 

1979 Personal communication. Landowner, Berkeley County, South 
Carolina. 

Piatt, William P. 

1979 Personal communication. Landowner, Berkeley County, South 
Carolina. 

Plog, Fred and Daniel G. Bates 

1976 Cultural anthropology . Alfred Knopf, New York. 

Polhemus, Richard 

1971 Excavations at Fort Moore-Savano Town (38Ak445). The Institute of 
Archaeology and Anthropology Notebook 111(6): 132-13TI University 
of South Carolina, Columbia. 



364 



Porcher, Frederick A., Esq. 

1852 Historical and social sketch of Craven County, South Carolina. In 
A contribution to the history of the Huguenots of South Carolina , 
compiled by T. Gail lard Thomas, M.D., pp. 87-168. R. L. Bryan, 
Columbia, South Carolina. 

Porcher, Thomas 

1822- Diary. South Carolina Historical Society. Charleston. 
1823 

Price, Cynthia R. 

1979 19th Century ceramics in the eastern Ozark border region. Center 
for Archaeological Research, Southwest Missouri University, Mono- 
graph Series, #1 , Springfield. 

Quimby, Ian 

1973 Cerami cs i n Ameri ca . The University Press of Virginia, 
Charlottesville. 

Radford, Albert E., Harry E. Ahles, and C. Ritchie Bell 

1968 Manual of the Vascular Flora of the Carol inas . University of 
North Carolina, Chapel Hill. ~ 

Rasmus sen, Wayne David (editor) 

1975 Agriculture in the United States: A documentary history (4 
vols.). Random House, New York. 

Ravenel , Henry 

1760- Ledger. South Carolina Historical Society, Charleston. 
1774 

1763- Day Book. South Carolina Historical Society, Charleston. 
1766 

Ravenel , Thomas Porcher 

1731- Family Papers. South Carolina Historical Society, Charleston. 
1906 

1829- Fragment. "Records due . . . from Negroes for Articles 
1833 Purchased". 

1809- Day Book of Henry Ravenel, Jr. 
1820 

1830- Crop Book for Planting 
1832 

1771- Checkbook of Slaves. 
1850 

1809- Checklist of Slaves. 
1867 

1832 Census of Pineville. 



365 



Rawick, George P. (editor) 

1972 The American Slave: A Composite Autobiography. Vol. 2 South 
Carolina Narratives. Parts 1 and 2. Contributions in Afro-Ameri- 
can and African Studies No. 11. Greenwood Publishing, Westport, 
Connecticut. 

Redfield, Robert, Ralph Linton and Melville J. Herskovits 

1936 Memorandum for the study of acculturation. American 
Anthropologist 38:149-152. 

Register Mesne Conveyance 

1719- Deed Books. Office of the Register Mesne Conveyance, Charleston. 
Present 

Rice, Prudence 

1976 Rethinking the ware concept. American Antiquity 41:538-542. 

Richardson, Emma B. 

1942- Dr. Anthony Cordes and some of his descendants. South Carolina 
1943 Historical Ma gazine 43:133-155; 219-242; 44:17-42; 115-123; 
184-195. 

Rose, Willy Lee 

1964 Rehearsal for reconstruction: the Port Royal experiment . The 
Bobbs-Merrill Co., Inc. New York . 

Rouse, Irving 

1960 The classification of artifacts in archaeology. American Anti- 
quity 25:313-23. 

Sabloff, J. and R. Smith 

1969 The importance of both analytic and taxonomic classification in 
the type-variety system. American Antiguity 34:278-85. 

Sal wen, Bertrand and Sarah T. Bridges 

1974 The ceramics from the Weeksville excavations, Brooklyn, New York. 
Northeast Historical Archaeology 3(l):4-29. 

Schopmeyer, C. S. 

1974 Seeds of Woody Plants in the United States. United States Depart- 
ment of Agriculture, Agriculture Handbook No. 450 . 

Schuyler, Robert L. 

1974 Sandy Ground: Archaeological sampling in a Black community. The 
Conference . Historic Site Papers 7 : part 2:13-51. 

1980 Archaeological perspectives on ethnicity in America Baywood 
Publishing company inc., Farmington, N.Y. 

Secretary of State 

1671- Miscellaneous Records, Main Series. South Carolina Department of 

1973 Archives and History, Columbia. 



366 



Secretary of State, Mortgages, Charleston Series 
1736-1867 

Secretary of State, Bills of Sale 

Secretary of State, Account Statements 

Secretary of State 

1731- Land Grants, Colonial Series. South Carolina Department of 
1776 Archives and History, Columbia. 

Secretary of State 

1785- Marriage Settlements. South Carolina Department of Archives and 
1902 History, Columbia. 

Shepard, Anna 0. 

1948 Plumbate: a Mesoamerican trade ware . Carnegie Institution of 
Washington, D.C. 

1965 Ceramics for the archaeologist. Carnegie Institution of Washing- 
ton: 609 , Washington, D.C. 

Singleton, Teresa 

1979 Slaves and ex-slave sites in Coastal Georgia . Paper presented at 
the 12th annual meeting, Society for Historical Archaeology, 
Nashville. 

Sirmans, H. Eugene 

1966 Colonial South Carolina; a political history, 1663-1763 . Insti- 
tute of Early American History and Culture, the University of 
North Carolina Press, Chapel Hill, North Carolina. 

Smith, Jerome Irving 

1967 The decoration of Whieldon ware. The Connoisseur , July: 198-204. 

Smith, Michael E. 

1979 A further criticism of the type-variety system: the data cannot 
be used. American Antiquity 44:822-825. 

Smith, Richard E., Gordon Willey, and James Gifford 

1960 The type-variety method as a basis for the analysis of Maya 
pottery. American Antiquity 25:330-40. 

Smith, Samuel D. 

1976 An archaeological and historical assessment of the first 

Hermitage. Tennessee Division of Archaeology Research Series 2 . 
Tennessee Department of Conservation, Nashville. 

South Carolina Gazette 

1732- Microfilm edition, Emory University. 
1782 



367 



South, Stanley 

1973 Indian pottery taxonomy for the South Carolina coast. Institute 
of Archeology and Anthropology, University of South Carolina , 
Notebook 5(Z):b4-bb, Columbia/ 

1974 Palmetto parapets: exploratory archaeology at Fort Moultrie, 
South Carolina 38CH50. Institute of Archeology and Anthropology , 
Anthropological Studies Y. Columbia, South Carolina. 

1976 An archaeological survey of southeastern Coastal North Carolina. 
Institute of Archeology and Anthropology , The Notebook VIII . 
Columbia, South Carolina. 

1977a Method and theory in histor ical archeology . Academic Press, New 
YorT. 

1977b Research strategies in historical archeology. Academic Press, New 
York": 

1977c An archeological survey of the southeastern North Carolina coast. 
Institute of Archeology and Anthropology, University of South 
Carolina, Notebook 8(1,2). Columbia. 

Spiro, Mel ford E. 

1955 The acculturation of American ethnic groups. American 
Anthropologist 57:1240-1252. 

Stampp, Kenneth M. 

1956 The peculiar institution; slavery in the ante-bellum South . 
Vintage Books, New York. 

Steen, Ivan D. 

1970 Charleston 1n the 1850' s: as described by British travelers. 
South Carolina Historical Magazine 71:36-45. 

Stoltman, James B. 

1974 Groton Plantation: an archaeological study of a South Carolina 
locality. Monographs of the Peabody Museum 1 . Harvard 
University, Cambridge. 

Stone, Lyle M. 

1970 Gun flints from eighteenth century Fort Michil imackinac. The 
Conference on Historic Sites Archaeology Papers , vol 5: 1-337" 

1974 Fort Michil imackinac, 1715-1781: An Archaeological Perspective on 
the Revolutionary Frontier. Publications of the Museum, Michigan 
State University , Anthropological Series 2, Lansing. 

Stoney, Samuel Gaillard 

1938 Plantations of the Carolina Low Country . Carolina Art Assoc. 
Charleston. 



368 



Storey, C. E. 

n.d. Trade Beads and other relics of historical interest found at Fort 
Moore. Ms. on file at the Augusta-Richmond County Museum, 
Augusta, Georgia. 

Surveyor General 

1731- Colonial Plats, Duplicates. South Carolina Department of 
1775 Archives and History, Columbia. 

Cordes, John 

1753 Vol. 5, p. 280. 

Cordes, Samuel 

1770 Vol. 11, p. 348. 

Cordes, Thomas 

1741 Vol. 2, p. 529. 

Cordes, Thomas 

1740 Vol. 12, p. 37. 

Sutherland, Donald 

1974 Excavations at Spanish Mount shell midden, Edisto Island, South 
Carolina. South Carolina Antiquities 6:25-36. 

Sussman, Lynne 

1977 Changes in Pearl ware dinnerware, 1780-1830. Historical Ar- 
chaeology 11:105-111. 

Terry, George D. 

1981 "The champagne country lies chiefly on the river"; a social 
history of an eighteenth century lowcountry parish in South 
Carolina ". Unpublished Ph.D. dissertation, University of South 
Carol ina. 

Thomas, John 

1887 Diary. 2 July 1845-8 April 1878 

Tindall, George 

1967 The emergence of the new South . Louisiana State University Press, 
Baton Rouge, Louisiana. 

U.S. Census Office, Washington, D.C. 

1840 Sixth Census of the United States, 1840. Microfilm of Enumera- 
tors' Manuscript Reports, South Carolina. Reel 1, Volume 2. 

1850 Seventh Census of the United States, 1850. Microfilm of Enumera- 
tors' Manuscript Reports, South Carolina. Reel 1, Volume 2, Reel 
5, Volume 4 (Slave Census). 

1860 Eighth Census of the United States, 1860. Microfilm of Enumera- 
tors' Manuscript Reports, South Carolina. Reel 1, Volume 2, Reel 
5, Volume 2 (Slave Schedules). 



369 



Ver Steeg, Clarence L. 

1975 Origins of a southern mosaic; studies of early Carolina and Geor- 
gTa i University of Georgia Press, Athens, Georgia. 

Vescelius, Gary S. 

1977 Newsletter of the Virgin Islands Bureau of Libraries, Museums and 
Archaeological Services , Information 2(3 ): 1-3. 

Vlach, John 

1977 Shotgun houses. Natural History , February: 51 -57. 

Waddell, Gene 

1980 Indians of the South Carolina Lowcountry, 1562-1751 . The Reprint 
Company, Publishers, Spartanburg. 

1981 Personal communication. 

Walker, Ian C. 

1967 Clay pipes from the fortress of Louisbourg Nova Scotia, Canada. 
The Conference on Historic Sites Archaeology Papers 1:96-99. 

Wallace, David D. 

1966 South Carolina: a short history . University of South Carolina, 
Columbia. 

Watson, Patty Jo 

1976 In pursuit of prehistoric subsistence: a comparative account of 
some contemporary floatation techniques. Midcontinental Journal 
of Archaeology 1:77-100. 

Weingarten, Lucille 

1979 Personal correspondence, editor Just Buttons Magazine . 

Whartenby, Franklee Gilbert 

1977 Land and labor productivity in United States cotton production . 
Arno Press, New York. 

Wheaton, Thomas R. , Jr. 

1976 La ceramica clasica del area de Huejotzingo, Puebla. 
Comunicaciones . 13:25-32. 

1980 Architecture at Curriboo and Yaughan Plantations . Paper presented 
before the Annual Meeting of the Southwestern Archeol ogical 
Conference, New Orleans. 

White, Wesley 

n.d. Report of historical research of the proposed Cooper River redi- 
version canal. Ms. on file at the Institute of Archeology and 
Anthropology, Columbia. 

Williams, Stephen (ed.) 

1968 The Waring papers: the collected works of Antonio J. Waring, Jr. 
Papers of the Peabody Museum of Archeology and Ethnology 58 . 
Cambridge. 



370 



Williams, Stephen and James B. Stol tman 

1965 An outline of southeastern United States prehistory with particu- 
lar emphasis on the Paleo-Indian era. In The Quaternary of the 
United States , edited by H. E. Wright and David G. Frey, pp. 
669-683, Princeton University, Princeton. 

Wilson, Eugene M. 

1979 Northern French construction in American folk houses . Paper 
presented at the Association of American Geographers 1979, 
Philadelphia. 

Wilson, Jack H. 

1979 European contact and plant food subsistence among the Carolina and 
Virginia Siouans . Paper presented at the 36th Annual Meeting of 
the Southeastern Archaeological Conference. 

Winberry, John J. 

1979 Reputation of Carolina indigo. South Carolina Historical Magazine 
80:242-250. 

Wood, Peter H. 

1974 Black majority; Negroes in colonial South Carolina from 1670 
through the Stono Rebellion^ W. W. Norton, New York. 

Woodward, C. Vann 

1951 Origins of the new South . Louisiana State University Press, Baton 
Rouge, Louisiana. 

Wright, Gavin 

1970 'Economic Democracy' and the concentration of agricultural wealth 
in the cotton South, 1850-1860. Agricultural History 64:63-93. 

Yarnell, Richard A. 

1974 Plant food and cultivation of the Salts Cavers. In Archaeology of 
the Mammoth Cave Area , edited by Patty Jo Watson. Academic Press, 
New York, pp. 113-122. 

Zaslavsky, Claudia 

1974 Africa counts. Prindle, Weber and Schmidt, Inc., Boston. 



APPENDIX A 
PREHISTORIC OVERVIEW 



A- 2 



This appendix addresses the prehistory of the project area. Anderson (1977) 
discussed the status of research into coastal South Carolina prehistory as of 
1976, noting that there had been "a complete absence of systematic archaeolo- 
gical investigation in the state until the mid-1960's" (Anderson 1977:19). 
Since the mid- '60s, he continued, research interest had quickened in this 
area, but as of the present, no regional synthesis exists. The following 
overview, therefore, relies upon the compilation by Garrow and Williams 
(1980) of information from intensive survey reports in coastal South Carolina 
as well as on the Cooper River Rediversion survey of the project area 
(Brockington 1980). The latter observes that while several studies address 
the South Carolina coastal plan, "detailed understanding of the ecological 
adaptations of human groups in the interior coastal plain awaits further 
investigation" (Brockington 1980:17). 

Paleo-Indian 

The Paleo-Indian Period, 12,000-8,000 B.C., is the earliest well -documented 
period of human occupation in this area, although an earlier occupation of 
North America has been hypothesized. Current research describes these people 
as nomadic groups of 25-30 members who followed the movement of game, particu- 
larly the now extinct megafauna. It is also believed that these groups also 
exploited wild plants, timing their movements to coincide with the availabili- 
ty of these plants (Brockington 1980:17). 

In the Great Plains and Southwest, sites of this period generally contain 
highly specialized projectile point types often associated with butchered re- 
mains of now extinct Pleistocene megafauna. Diagnostic Paleo-Indian arti- 
facts, such as lanceolate (sometimes fluted) projectile points, small unifa- 
cial end scrapers, knives, drills, and graving tools have been found through- 
out the eastern United States (Griffin 1967), but no associations with ex- 
tinct fauna have been recorded there (Garrow and Williams 1980:46). Fluted 
point data from South Carolina suggest that Paleo-Indian points occur most 
frequently on the terraces of major coastal plain drainages (Michie 1967). A 
distributional study of fluted points suggests that Paleo-Indian points tend 
to be concentrated in the Interior Low Plateau of North America along major 
river drainages (Williams and Stoltman 1965:676). This is attributed to 
changes in sea levels; sites of this period may be under water or deeply 
buried beneath alluvial deposits (Garrow and Williams 1980:47). It is, none- 
theless, believed that human occupation was not dense in this period, and no 
Paleo-Indian sites were discovered during the survey of this project area 
(Brockington 1980:17). An unfluted lanceolate point was reported, however, 
near the site of the Santee-Cooper Canal in nearby St. John's Parish, Berke- 
ley; and while digging the original Santee Canal in the 1790s, Pleistocene 
faunal remains were found east of the Santee-Cooper plant site at a depth of 
nine feet (Herold and Knick 1978:10). 

Archaic 

The Archaic Period, extending from about 8,000 to about 1,000 B.C., is seen 
as a long period of human re-adaptation to a changing environment. Conven- 
tionally, it has been sub-divided into three phases: Early Archaic (8,000- 
6,000 B.C.), Middle Archaic (6,000-4,000 B.C.), and Late Archiac (4,000-1,000 
B.C.). Not all of these have been documented for the coastal area. Prior to 



A-3 



3,000 B.C., the shoreline lay east of its present location and, therefore, 
sites formed before that time are under the Atlantic Ocean (DePratter 1975: 
10). Nomadic hunting and gathering by small groups, with increasing emphasis 
on wild plant and fish resources are characteristic for the period as a 
whole. Caldwell (1958) interprets this period as a trend toward "Primary 
Forest Efficiency", in which groups became more familiar with the resources 
of the eastern forests. Population grew, and toward the end of the period, 
political and economic organization may have become more complex (Brockington 
1980:18). Changes in point types form the basis for differentiating among 
the phases of the Archiac Period. 

The Early Archaic phase is seen as a succession of responses to post-Pleisto- 
cene conditions. The subsistence patterns have been characterized as hunting 
and trapping a wider variety of animals, increased reliance on riverine re- 
sources, and increasing development of seasonal food gathering (Griffin 
1967). The tool assemblages, however, remain unchanged for the most part, 
from the Paleo-Indian Period, but projectile point forms change from lanceo- 
late to stemmed or notched. Ground stone artifacts and grinding implements 
are rare or missing (DePratter 1975:4). Early Archaic point types have been 
systematically recovered primarily in Piedmont North Carolina, Tennessee and 
West Virginia, although a Palmer corner notched point and a possible Palmer 
variant were found at the Cal Smoak site on the upper Edisto River in the 
coastal plain of South Carolina. Point types that may be earlier, transi- 
tional or contemporaneous with assemblages at other sites in the upper Edisto 
River region were also reported in the Atlantic Intracoastal Waterway study 
(Garrow and Williams 1980:47-48). 

The Middle Archaic phase is characterized by a wider variety of ecological 
niches and a seasonally-based hunting and gathering subsistence strategy 
(Caldwell 1965). Diagnostic tools include stemmed bi facially worked projec- 
tile points and atlatl weights, as well as a marked increase in ground tools 
for plant processing. 

Middle Archaic artifacts have been recovered through Coastal and Piedmont 
South Carolina and at the sites discussed have. This has suggested a distri- 
butional pattern on swamp margins, particularly on terraces overlooking flood- 
plains, marked by decreasing artifact density with increasing distance from 
optimum settlement zones. Garrow and Williams (1980) recorded two isolated 
Middle Archaic finds in their survey of the Atlantic Intracoastal Waterways 
(AIWW) in the tidewater east of the project area. A Guilford projectile 
point was recovered by a local informant near his home on the Waccamaw River, 
which is somewhat north of the project area. Garrow and Williams 
(1980:48-49) also reported a Morrow Mountain II Middle Archaic projectile 
point, recovered on the eroding beach of Coosaw Island. This had been 
redeposited by tidal action, and no additional prehistoric arti factual 
material was noted. 

The Late Archaic phase is one of the better documented periods in the cultur- 
al sequence of coastal South Carolina. It is characterized by the appearance 
of shell mounds and shell rings, and, significantly, by the introduction of 
ceramics. It also indicates the development of a more sedentary pattern and 



A- 4 



more complex social, economic, and religious organization (Brockington 1980: 
18). Although the survey of the Cooper River Rediversion Canal area did not 
recover artifacts from this period, Garrow and Williams encountered Late Ar- 
chaic sites frequently in their survey nearer the coast. These were, how- 
ever, "predominantly isolated finds or severely eroded artifact scatters" 
(Garrow and Williams 1980:49). 

In coastal South Carolina, Late Archaic sites show Stall ings and Thorn's Creek 
ceramics in addition to Savannah River stemmed, Thelma, and Gary projectile 
points, and a wide range of lithic materials (Anderson, Lee, and Parler 1979: 
92). The occurrence of Stall ings and Thorn's Creek ceramics have generated 
concern over temporal priorities. In the Savannah River basin, Stall ings 
ceramics were found to be stratigraphically earlier than Thorn's Creek (Stolt- 
man 1974:86). On the Edisto River, however, Sutherland (1979) found Thorn's 
Creek to have greater antiquity at Spanish Mount. Though Stall ings ceramics 
still hold the earliest radio-carbon dates (Stoltman 1974), investigators now 
tentatively agree that the two ceramic groups are coeval. Additionally, it 
is believed that the Awendaw Punctate, initially identified as a variant of 
Thorn's Creek, but now considered a third ceramic type, dates from a time 
prior to Thorn's Creek and Stall ings Island on the basis of a radiocarbon date 
from an oyster shell. Corroborating evidence has not yet been recovered to 
substantiate the radiocarbon data (Garrow and Williams 1980:52). 

These three ceramic types have been discovered to have a clear distributional 
pattern throughout the Coastal Plain of South Carolina and Georgia. Fiber 
tempered wares are popular in the southwest part of the Coastal Plain of 
South Carolina between the Edisto and Savannah Rivers; they predominate in 
the Georgia Coastal Plain. Thorn's Creek ceramics are absent in Georgia but 
are widespread in South Carolina, concentrating in the region between the 
Edisto and Pee Dee Rivers. It has been hypothesized that the region between 
the Edisto and Savannah Rivers, where both types occur, represents an area 
where the different cultural groups overlap (Garrow and Williams 1980:52). 

The temporal organization of the data is made more complicated by other 
trends evident in the artifacts. In addition to variations in ceramic styles, 
investigators have shown variations in cultural assemblages. Bull en and 
Greene (1970) point out a change in projectile point morphology from large 
rhyolite forms in the preceramic levels to smaller quartz and flint points in 
the upper levels of Late Archaic sites. The manufacture of steatite vessels 
is thought to have begun during the pre-ceramic phase (Stoltman 1972; Griffin 
1952). The occurrence of steatite vessel fragments is relatively common 
throughout the Piedmont Eastern United States, but the transition from stea- 
tite to ceramic vessels resulted in a division in choice of tempering. Along 
the Coastal Plain/Piedmont from New York through the Carol inas, the earliest 
pottery has steatite particles (Griffen 1952); south of these areas, the ear- 
liest pottery has been vegetable tempered. In 1960, both steatite and fiber- 
tempered pottery sherds were recovered in an archaeological survey of the 
southeastern North Carolina and northeastern South Carolina coastal area 
(South 1976), but no steatite-tempered pottery was recovered along the South 
Carolina coast in the AIWW survey in 1978-1979, although the investigators do 
not rule out the potential for discovering its presence there (Garrow and 
Williams 1980:53). 



A-5 



Finally, a Late Archaic model of temporary base camps in which base camps 
were located near or within coastal marshes or river swamps in order to uti- 
lize riverine shellfish resource, has been suggested. Specialized extraction 
areas were located in the uplands or internal Coastal Plain, and the large 
shell middens and rings on the coast have been interpreted as focal or pri- 
mary occupation areas (Stoltman 1972:50). Not surprisingly, since its invest- 
igation was confined largely to the sea is! and/ tidewater zone, the AIWW sur- 
vey recovered information indicating the importance of shellfish in the diet 
(Garrow and Williams 1980:53). 

Investigation at Cal Smoak and other Late Archaic sites in the interior Coast- 
al Plain suggest a somewhat different interpretation. Anderson noted the 
lack of shell middens at many interior Coastal Plains sites, as well as the 
diversity of faunal and floral remains at sea island sites of this period, 
and suggested that subsistence activities continued to follow the exploita- 
tion of white-tailed deer and plant resources. On the basis of artifact dis- 
tribution a more sophisticated social organization composed of endogamous, 
possibly tribal level groups are hypothesized (Anderson, Lee, and Parler 
1979:93-95). 

Woodland Period 

The Woodland Period dates from about 1000 B.C. to A.D. 1000 and is repre- 
sented at 38BK76. The earlier pattern of subsistence persists; manufacture 
of ceramics becomes more widespread; first evidence of systematic horticul- 
ture becomes evident, and burial mound construction appears. Group size is 
believed to have increased, and there was a decided trend toward sedentism. 
Migration patterns seem to have tended toward scheduled seasonal rounds of 
group movements, believed to involve segmentation and re-merging of groups in 
a cyclical fashion. The system derived from an effort to optimize the still- 
dominant hunting/fishing/wild plant gathering subsistence focus (Brockington 
1980:18). 

Three ware groups are associated with this period: Refuge, Deptford, and 
Wilmington-Cape Fear. Refuge ceramics are known from two excavated sites, 
Refuge (Williams 1968) and Groton Plantation (Petersen 1971), and from scat- 
tered surface occurrences along the South Carolina coast as far north as the 
PeeDee River. Radiocarbon dates place them between 800 and 1200 B.C. (Peter- 
sen 1971:77, 80), and they are believed to have developed out of Stall ings or 
Thorn's Creek ceramics (Williams 1968). This ceramic type is also believed to 
be typical of a subsistence pattern that shared the earlier pattern based on 
shellfish and white-tailed deer associated with the Late Archaic phase (De- 
Pratter 1977:11). Refuge phase sites, however, were also found outside the 
swamp-river system not in association with the shell middens common at Stal- 
lings and Thorn's Creek occupations. The AIWW survey did not find Refuge type 
sites, which also suggests a non-marine focus in the subsistence pattern. It 
is believed that the Refuge phase subsistence base was hunting and gathering 
rather than shellfish or agriculture, although resolution of this question 
awaits further information from controlled excavation. It is presently held 
that Refuge helps bridge the gap between Stall ings/Thom's Creek and the later 
Deptford phases (Garrow and Williams 1980:55). Refuge ware was recovered 



A- 6 



during the survey of the Santee-Cooper Power Plant site in St. John's, Berke- 
ly, Parish (Herold and Knick 1978:26). Although Anderson found a high inci- 
dence of Thorn's Creek and Deptford wares in association with Refuge ware, no 
Thorn's Creek-like pottery was recovered at this Berkeley County site although 
several sherds suggested the presence of Deptford ware (Herold and Knick, op 
cit ) . 

Deptford ceramics follow the Refuge ware. They are widespread throughout 
coastal Georgia and southern South Carolina as far north as Charleston (South 
1976:1-2). From the vicinity of Charleston north to the vicinity of Cape 
Fear, North Carolina, the ceramics are dominated by a sherd tempered ware 
with fabric impressed finish that differed from Deptford surface designs, 
which are mainly paddle and check -stamped. These ceramics have been termed 
the Hanover series (South 1976:1-2). South has placed Hanover ceramics in 
the Wilmington Ware group, signifying an early component of this type, rough- 
ly contemporary with Deptford. In South Carolina in general, Wilmington re- 
mained entrenched in the coastal area until well into the Mississippian peri- 
od at about A.D. 1100 (South 1973). In the north coastal area, the Hanover 
Series Wilmington Ware group pottery was replaced by Cape Fear sand tempered 
cord and fabric decorated ware (South 1976:1-2). At a later time, Cape Fear 
sand tempered fine cord marked extended to the Savannah River, where it is 
known as Savannah Fine Cordmarked, a pottery type within South 's (1973) Cape 
Fear Ware group (South 1976:2). 

The different surface decorations are construed to reflect ethnic variation 
within the underlying population by most southeastern archaeologists. Find- 
ing both kinds of pottery in South Carolina has been thought to indicate 
either the intermingling of separate peoples or the stylistic influence from 
separate culture centers. Both Cape Fear-Wilmington and Deptford sites were 
found in the project area in the Cooper River Rediversion survey (Brockington 
1980:12-19). The Wilmington ceramic type has been found largely in the area 
south of Charleston and north of the Altamaha River in Georgia. The AIWW 
survey recovered evidence of Wilmington occupation in upland areas, and their 
recovery of a sherd tempered heavy cord marked Wilmington potsherd on the 
Waccamaw River, about 20 miles north of Georgetown, which is substantially 
north and east of the present project site, "is possibly the furthest north 
that a Wilmington component has been recovered and documented" (Garrow and 
Williams 1980:56). Herold and Knick (1978:26) also found instances of the 
Cape Fear-Wilmington ware groups in the survey area for the site of the San- 
tee-Cooper Power Plant in Berkeley County, which is west of the present pro- 
ject area; their preliminary findings suggest that Wilmington may have pre- 
dominated over Cape Fear. 

Mississippian Period 

The Mississippian Period followed the Woodland Period, dating from about A.D. 
1000 to about 1550, the period of European contact. This period may be cha- 
racterized by intensive agriculture in some areas, sedentary populations, 
truncated temple mounds, and the development of a highly stratified socio- 
political organization (Garrow and Williams 1980:57). The changes associated 
with this period are said to be the consequence of the impact of ideas from a 
Mississippian center located outside this region on the indigenous culture, 
although this interpretation has tended to exaggerate the differences between 



A- 7 



Mississippi an and pre-Mississippian periods (Hally 1975:47). As indicated 
earlier in the discussion of the distribution of Cape Fear-Wilmington ceram- 
ics along the coast of South Carolina, there was at least some continuity 
between the Woodland and Mississippi an Periods in this area. The idea of a 
distinctive, regional Mississippian tradition called South Appalachian Missis- 
sippian, moreover, was first proposed by Holmes in 1903 (cited in Anderson 
1977:9) and adopted subsequently by Griffin (1967) and Ferguson (1971). At- 
tributes of the South Appalachian province include stamped pottery, burial 
mounds and stratified society as reflected in burials, village sites and 
shell heaps, location adjacent to large water courses and in the larger com- 
plexes, dependence on the cultivation of maize. 

Much emphasis has been placed on the cultivation of maize as a prerequisite 
for a Mississippian way of life. Historical sources of the period, however, 
indicate that even at European contact, Indians still depended upon marine 
resources and hunting and gathering for subsistence. Reference is made to 
corn in historic accounts, but it has been suggested that there was less de- 
pendence on agriculture in coastal South Carolina than in areas removed from 
the coast as a result of the availability of alternative marine foodstuffs 
(Garrow and Williams 1980:58). 

In addition to the persistence of cord- and fabric-impressed Cape Fear ceram- 
ics (South 1977a:2), several other ceramic types have been identified for the 
Mississippian Period in South Carolina. These include the Charles Town, Mul- 
berry, Fort Watson, Adamson, Pee Dee, Irene, and Savannah types. South (1973) 
groups these under the all -encompassing ware group termed Chicora, which in- 
cludes the sand tempered complicated stamped, burnished pottery, often with 
rosettes, reed punctations, and punctated rim strips (South 1976:28) charac- 
teristic of Mississippian sites. 

In his report of the survey of the Cooper River Redi version site, Brockington 
noted the importance of this area to Mississippian development in this re- 
gion. He observed, however, that many Mississippian sites had been destroyed 
without investigation 1n the creation of Lakes Marion and Moultrie (Brocking- 
ton 1980:19). 

Prehistoric Artifacts 

The prehistoric components at the sites were, with one exception, sparse and 
scattered. A range of Middle Archaic through Mississippian artifacts were 
found and are listed below by site. Sites 38BK75, 38BK76, and 388K245 had 
276, 1491, and 24 prehistoric artifacts respectively. 

The only diagnostic lithic artifacts were a Morrow-Mountain II point of fos- 
sil iferous chert and a mid to late Archaic proximal fragment with a stem from 
the surface at 38BK75; and a Morrow-Mountain II of white chert from Feature 
F4 at 38BK245. A slate gorget fragment was also found in Feature F12 at 
38BK245. Diagnostic material found in historic trash features may show 
curation of such artifacts by the historic inhabitants, however. 

Ceramics were more diagnostic and run the gamut from early Woodland to Missis- 
sippian. No analysis beyond description of the prehistoric artifacts was at- 
tempted for three reasons. 



A-8 



1. Many prehistoric sites containing structures, burials, hearths, and other 
features were excavated during the project by Commonwealth and Associates and 
by the Institute of Archeology and Anthropology. 

2. The material at Yaughan and Curriboo was thin and widely scattered, ex- 
cept around Structure 76B, and much of the material came from the surface. 

3. No prehistoric features, either postholes or middens, were located des- 
pite extensive excavation and mechanical stripping. 

The material will be curated and available for future analysis at the Insti- 
tute of Archeology and Anthropology (IAA) in Columbia, South Carolina. 

The material found around Structure 76B is listed separately. This material 
represents the largest concentration found during the project and is undoubt- 
edly an extension of Site 38BK236, which was excavated by IAA. Sixty-eight 
percent of the material at Structure 76B was either from the surface or root 
mat levels. If material from historic features is added to that, the amount 
of disturbed artifacts reaches a minimum of 75 percent. This high rate of 
disturbance coupled with a lack of features makes even Structure 76B an un- 
likely candidate for significant insight into prehistoric settlement/sub- 
sistence patterns. 



TABLE A-l. Prehistoric Materials 
Recovered from Historic Sites 









Structure 76B 


(only) by 


layers 








Dark 




Total 


38BK- 


38BK- 


38BK- 


Root Grey 


Historic 


Structure 


75 


76 


245 


Surface Mat Sand 


Features 


Baulks 76B 



A- 9 



Angular shatter 


19 


194 


1 


13 


77 


30 


13 


18 


151 


Biface thinning flake 


162 


454 


4 


10 


220 


110 


34 


39 


413 


Decortifi cation flake 


7 


122 


2 


6 


69 


22 


13 


7 


117 


Core 


1 


24 


1 


2 


16 


1 




1 


20 


Biface (or fragment) 


24 


25 


2 


2 


5 




2 


1 


10 


Thumbnail scraper 


1 


1 


1 




1 








1 


Gorget 






1 














Mano 




9 






1 








1 


Cobbl es 


4 


26 






14 






1 


15 


Utilized Flakes 




12 






4 








4 


Abraders/poss abraders 


19 


4 


2 














Fire cracked rock 


1 


14 












2 


2 


End scraper 


1 


1 
















Uniface 


1 


5 




1 


1 


1 


1 




4 


Celt 




1 
















Biface blank 




3 










1 




1 


Other/uni denti f i ed 




6 


1 




5 






1 


6 


Total lithics 


240 


901 


15 


34 


413 


164 


64 


70 


745 


Check stamped 




93 




24 


53 


3 


4 


8 


92 


Fabric impressed 


25 


107 




23 


72 


8 


4 


7 


114 


Incised 




1 


1 














Cord marked 




2 
















Complicated stamped 


1 


4 


4 














Plain 


10 


352 


3 


62 


229 


36 


11 


40 


378 


Linear check stamp 




2 


1 


2 


2 








4 


Sherd tempered 




10 






8 


1 




1 


10 


Fiber tempered 




3 




1 


1 






1 


3 


Rou letted 




9 






5 




3 




8 


Other decoration 




7 




1 


4 




1 




6 


Total ceramics 


36 


590 


9 


113 


374 


48 


23 


57 


615 


TOTAL 


276 


1491 


24 


147 


787 


212 


87 


127 


1360 



APPENDIX B 
OWNERSHIP OF THE PLANTATIONS 



B-2 



Yaughan Chain of Title 

The earliest known reference to Yaughan Plantation is a document describing a 
transaction in 1737 between Richard Allein of St. James, Santee, Parish and 
Edward Thomas of the same parish (Indenture, 27 July 1737, SCHS). It con- 
cerns a tract of land measuring 650 acres "known by the name of Yaughan," 
which was part of a larger 1200-acre tract, located in the Parish of St. 
James, Santee, bordering on lands of Thomas Ellery (Figure 19). The document 
is fragmentary, and therefore, it is impossible to decipher all of it. Later 
transactions, however, allude to the Allein/Thomas exchange and confirm its 
content. 

Yaughan Plantation was then conveyed by Edward Thomas to his son Samuel, who 
in turn, sold two tracts of land to Isaac Cordes in 1742. One was a 596-acre 
tract in St. James, Santee, bordering on the Santee River, transferred from 
Edward Thomas to Samuel in 1740 (Indenture, 27 February 1740, Deed Book R-5, 
p. 186, RMC). The second was an adjoining 650-acre tract "being part of a 
larger Tract of Twelve hundred acres of Land. . .commonly called and known by 
the name of Yaughan" (Samuel Thomas and his wife to Isaac Cordes, 26 January 
1742, Deed Book R-5, p. 188, RMC). Isaac Cordes acquired both tracts in the 
1742 transaction (Samuel Thomas and his wife to Isaac Cordes, 26 January 
1742, Deed Book R-5, p. 188, RMC). 

Isaac Cordes died in 1745 (S.C. Gazette, 13 July 1745). In his will, re- 
corded 9 August 1745, he left his plantation in the Parish of St. John, 
Berkeley, to his only son together with "all other my Lands and real Estate 
whatsoever" (Will of Isaac Cordes, recorded 9 August 1745, Record of Wills, 
Vol. 5, 1740-1745, p. 407, CCPL). Land was not appraised in colonial South 
Carolina as part of probating a will (Main 1965:64), but the inventory of the 
estate of Isaac Cordes, taken 9 August 1745, lists cattle, oxen, sheep, hogs, 
horses and some household goods at "Youshan" (Inventory of Isaac Cordes, 9 
August 1745, Inventories, Vol 67, Book A, 1732-1746, p. 329, CCPL). Since 
Isaac Cordes had clearly acquired the plantation three years earlier, this is 
evidently a mi spelling of the name "Yaughan." 

John Cordes died in 1756, and his will indicated "all my real Estate to be 
Equally divided between them," meaning his two sons John (born 1749) and 
Thomas (born 1753). He stipulated that his estate was to remain intact with 
the proceeds used to defray the expenses of supporting his wife and five 
children and educating them. The estate would gradually be partitioned at 
the marriages of his daughters and when his "Eldest Son shall Attain the Age 
of twenty one Years" (Will of John Cordes, 20 June 1756, recorded 3 December 
1756, Record of Wills, Vol. 7, 1752-1756, pp. 582-583, CCPL). His elder son 
John (born 1749) received his share of his father's estate in December 1768 
(Memorandum, 26 December 1768, Account Book, Estate John Cordes, 1764-1798, 
p. 69, CO. This evidently included Yaughan. Local tradition claims that 
John inherited Yaughan by law of primogeniture, which stated that the eldest 
son acquired all, or a major part of the patrimony, by right of birth (DuBose 
1852:50-51). John Cordes (1718-1756), however, clearly left his real estate 
evenly divided between his two sons; the ambiguity consisted in what lands he 
owned and which son would inherit which tract. 



B-3 



In addition to a plantation in Berkeley County and to the two tracts (includ- 
ing Yaughan) in Craven County, John (1718-1756) acquired a 1000-acre tract in 
St. James, Santee, surrounded on all sides by "vacant land" (Royal Grant to 
John Cordes, 2 March 1752, Royal Grants, Vol. 5, p. 130; Plat for John 
Cordes, 2 March 1753, Colonial Plats, Vol. 5, p. 280, SCDAH). John Cordes 
(1749-1798) sold three tracts to his younger brother Thomas (b. 1753) in 
1775, consisting of the two tracts that their grandfather Isaac had acquired 
from Samuel Thomas and bequeathed to their father and a third tract of 500 
acres, which was one-half of the 1000-acre grant their father had obtained in 
1753. The deed details the transaction thus: 

All that Plantation or Tract of Land containing Six hundred 
and Fifty Acres situate lying and being in St. Stephen's 
Parish. . .commonly called and known by the name of Yaughan. . 
.Also all that other Plantation or Tract of Land containing 
Five hundred and ninety six acres situate lying & being in St. 
Stephen's Parish aforesaid Butting and bounding Northwesterly 
on the aforesaid Tract of Six hundred and fifty acres. . . And 
also all that other Plantation or Tract of Land Containing 
Five hundred acres situate lying and being in the said Parish 
being the Northwestermost half part of a larger Tract of One 
thousand acres of Land Originally Granted to John Cordes 
deceased Father of the said John Cordes Party to these 
presents. . . (John Cordes to Thomas Cordes, Release of 3 
Tracts of Land, 10 May 1775, Deed Book R-5, pp. 194-195, RMC). 

The will of Thomas Cordes (1753-1806) is not extant, but his widow Charlotte 
Evance Cordes left "my Property equally among my Children" when she died in 
1826 (Will of Charlotte Cordes, 12 June 1826, recorded 27 February 1827, Vol. 
37, 1826-1834, p. 238, CCPL). The terms of Thomas Cordes' will may be in- 
ferred from her will and a tax return by the Executors of Thomas Cordes' 
Estate, St. Stephen's, 17 March 1825 (Comptroller General, Tax Return, St. 
Stephen's, 1824, SCDAH). In this return, Charlotte Cordes signed on behalf 
of an estate assessed taxes on a total of 1496 acres in the parish. Presum- 
ably, this included Yaughan. 

In 1836, three daughters of Thomas and Charlotte Cordes: Margaret Catharine, 
Charlotte Lavinia and Anna Camilla, sold their share of the real estate to 
Solomon Clarke (Title, 1 January 1836 / 23 January 1836, Deed Book M-10, pp. 
221-223, RMC). In January 1850, Clarke sold "all that Plantation or Tract of 
Land. . .known by the name of 'Yaughan,'" described further as bordering on 
Corriboo, to J. W. Thurston (Title, 1 January 1850, Deed Book Q-12, pp. 
75-76, RMC). Clarke died later that year (Will of Solomon Clarke, 16 October 
1850, recorded 11 November 1850, Record of Wills, Vol. 45, Book A, 1845-1851, 
pp. 771-773, SCDAH). By the mid-nineteenth century, the name "Yaughan" ap- 
plied to the river section as well; the estate of Solomon Clarke in 1852 sold 
880 acres in St. Stephen's, described as being on "Yaughan Branch" and bor- 
dering on the Santee River, to John W. Thurston (Title, 9 March 1852, Deed 
Book V-12, p. 619, RMC) 



B-4 



Thurston sold 182 acres to John Money in 1857, described as bordering "East 
by lands of J. W. Thurston," and located in St. Stephen's Parish (Thurston to 
Money, Conveyance, 1 October 1857, Deed Book W-13, pp. 475-476, RMC). Money 
evidently acquired more land from Thurston, because shortly thereafter he 
sold 800 acres "known as Yauhan" to Ann E. Piatt, which was further described 
as bounding "to the North on Santee River" (Title, 6 January 1858, Deed Book 
L-14, p. 244, RMC). Thurston retained part of the old Yaughan Plantation; in 
1863, a sale of land contained in Curriboo plantation positioned the parcel 
on "Lands of the estate of Thurston, formerly of Solomon Clarke called Yahan" 
(R. Press Smith to H. Panzerbeiter, Title, October 1862, Deed Book R-14, p. 
334, RMC). 

In 1869, Piatt directed that land consisting of "one half of the land con- 
tained in the two tracts. . .known as Harleston Hill and Yaughan plantation" 
be held in trust for William D. Piatt and J. C. Lequeux (Conveyance, 17 May 
1869, Deed Book H-15, pp. 548-552, RMC). Addison Lequeux became trustee for 
Piatt upon the death of S. D. Russell and in 1891, Lequeux, acting on behalf 
of Piatt, conveyed the "swamp lands of Yahaw [Yaughan] Plantation," which 
bordered to the east on Curriboo, to Samuel G. Stoney (Lequeux to Stoney, 
Indenture, 3 January 1891, Deed Book C-2, p. 765, Monck's Corner Assessor's 
Office; Title, 3 January 1891, Deed Book A-8, p. 10, Monck's Corner Asses- 
sor's Office). 

In 1906, Piatt sold the timber rights of "Yawhaw" to Y. Rittenberg and M. A. 
Floyd (Piatt to Rittenberg and Floyd, Deed 4 Contract, 16 May 1906, Deed Book 
C-7, p. 73, Monck's Corner Assessor's Office). After 1909, Piatt and his 
wife sold or gave 106-acre parcels, all part of the old Yaughan Plantation, 
to their three sons: Lewin Drake Piatt, Samuel J. Piatt and William P. Piatt 
(Personal Communication, 3 April 1979; Title, 30 October 1909, Deed Book 
A-31 , p. 182, Monck's Corner Assessor's Office; Title, 30 October 1909, Deed 
Book A-31, p. 183, Monck's Corner Assessor's Office). Samuel J. Piatt gave 
his 106-acre parcel to his nephew J. Lamar Piatt (Personal Communication, 3 
April 1979); Lewin D. Piatt sold his 106-acre parcel to J. R. Coggeshall and 
J. T. Coggeshall (Piatt to Coggeshall and Coggeshall, Title, 5 December 1912, 
Deed Book A-38, p. 158, Monck's Corner Assessor's Office) in 1912, and Wil- 
liam P. Piatt sold his 106-acre parcel to a party named Funk (Personal Commu- 
nication, 3 April 1979). In 1913, Coggeshall 4 Coggeshall sold the land to 
L. B. Browder (Coggeshall and Coggeshall to Browder, Title, 29 December 1913, 
Deed Book A-41 , p. 50, Monck's Corner Assessor's Office), who sold it to C. 
H. Browder in 1920 (Browder to Browder, Title, 13 February 1920, Deed Book 
A-49, p. 80, Monck's Corner Assessor's Office). In 1974, it was sold to Leo 
0. Browder, who then sold one-fourth interest to C. Harry Browder in 1975 
(Browder to Browder, Master's Title, 14 February 1974, Deed Book A-276, p. 
110, Monck's Corner Assessor's Office; Browder to Browder, Title, 22 October 
1975, Deed Book A-298, p. 27, Monck's Corner Assessor's Office). 

Curriboo Chain of Title 

A plat affixed to the Indenture conveying Yaughan from Richard Allein to Ed- 
ward Thomas indicates land to the southeast of Yaughan described as "Curriboo 
Land: the late John Moore Esq" (Figure 19). According to the 1737 Allein/Tho- 
mas transaction, Thomas Ellery had acquired the land belonging to the late 



B-5 



John Moore and thus appears to have owned Curriboo in 1737 (Indenture, 27 
July 1737, SCHS). Ellery died the following year but made no reference to 
Curriboo in his will. He left, however, the vast majority of his estate, 
both real and personal, to his wife Ann Ellery (Will of Thomas Ellery, 2 
October 1738, recorded 7 March 1738/9, Record of Wills, Vol. 4, 1736-1740, 
pp. 115-116, CCPL). 

When Yaughan was sold by Samuel Thomas to Isaac Cordes (d. 1745) in 1742, the 
description of the 650-acre tract referred to lands southeast of Yaughan "be- 
longing to Isaac and Thomas Cordes" (Samuel Thomas and his wife to Isaac 
Cordes, 26 January 1742, Deed Book R-5, p. 187, RMC). This is conceivably 
Curriboo, which had been acquired from Ellery or from his estate by Isaac 
Cordes and his brother Thomas, known as "Colonel" Thomas Cordes (Richardson 
1942(43):137), who had witnessed Ellery's will in 1738 (Will of Thomas El- 
lery, 2 October 1738, recorded 7 March 1738/9, Record of Wills, Vol. 4, 
1736-1740, p. 116, CCPL). When Isaac Cordes' personal property was inven- 
toried in August 1745, the appraisers listed slaves, livestock and "Sundries 
at Correboo between Coll [Thomas] Cordes and the Estate" (Inventory of Isaac 
Cordes, 9 August 1745, pp. 328-330, CCPL). By 1745, therefore, the Cordes 
brothers had evidently acquired Curriboo. 

Thomas Cordes died in 1748 and willed "all that my Plantation commonly called 
Correboo containing about One Thousand three hundred & ninety acres of land 
the same a little more or less scituate in Craven County" to his second son 
Samuel (d. 1796) (Will of Thomas Cordes, 25 April 1748, recorded 21 April 
1749, Record of Wills, Vol. 6, 1747-1752, p. 142, CCPL). Since inventories 
did not list real estate, Isaac Cordes may have had only a part interest in 
slaves and other personal property at Curriboo and no interest in the real 
estate itself. Alternatively, Thomas Cordes may have severed the relation- 
ship between Curriboo and Isaac Cordes' estate after his brother's death. 
Nonetheless, Curriboo was clearly in the possession of Thomas Cordes at his 
death in 1748. 

Samuel Cordes bequeathed his plantation in St. Stephen's Parish "known by the 
name of Curriboo" to his eldest son Thomas (d. 1799) in 1796 (Will of Samuel 
Cordes, 15 May 1796, recorded 29 October 1796, Record of Wills, Vol. 26, Book 
B, 1793-1800, p. 504, CCPL). Three years later, Thomas Cordes, also known as 
Thomas Cordes, Jr., to distinguish him from his cousin Thomas Cordes 
(1753-1806), who then resided at Yaughan, willed three plantations: 
"Curriboo Plantation with the Pine Land adjoining," "my Plantation and Lands 
in Saint Johns Parish Berkly County, together with the Plantation called 
Wiskinboo," to his only son James Jamieson Cordes (b. 1798) (Will of Thomas 
Cordes, Junior, 1 September 1799, recorded 20 March 1800, Record of Wills, 
Vol. 27, Book C, 1793-1800, p. 960, CCPL). 

In 1845, James Jamieson Cordes, John M. Harleston, T. C. Harlston and E. 
Harleston sold Charles Macbeth "All that Plantation or Tract of Land Called 
'Curriboo'. . .Bounding. . .on Lands belonging to Solomon Clarke, called 
Yahan," and consisting of 2255 acres (Title, 10 June 1845, Deed Book V-ll , p. 
45, RMC) (Figure 20). This transaction excluded five acres set aside "for a 
Burial Ground for St. Stephens Church" (Title, 10 June 1845, Deed Book, 



B-6 



V-ll:45, RMC). The obligation of Curriboo to St. Stephen's Parish was long- 
standing. In 1846, James J. Cordes, John M. Harleston, Thomas Cordes 
Harleston and Elizabeth C. Harleston conveyed four acres "on which the old 
Parish Church of St. Stephen stands" to the Parish of St. Stephen, citing a 
prior grant by the "then owner of the Plantation known as Curriboo" to the 
Parish of St. James, Santee, for the construction of a chapel of ease. When 
St. Stephen's Parish was created out of St. James, Santee, in 1754, the pro- 
posed chapel of ease became the parish church (T. C. Harleston, atty. for J. 
J. Cordes, et al . to Parish Church of St. Stephen's, Quit Claim, 16 June 
1846, Deed Book X-ll , p. 57, RMC). 

In 1849, Macbeth sold the entire 2255-acre tract "called 'Curriboo'" to 
Robert Press Smith (Macbeth to Smith, Conveyance, 16 May 1849, recorded 19 
February 1850, Deed Book C-12, pp. 538-539, RMC). The conveyance noted that 
the plantation included "a Tract of Pine land attached to the above described 
Plantation as will appear by the will of Thomas Cordes Junr deceased saving & 
reserving Five acres for a Buryal Ground for Saint Stephens Church. . . ." 
(Conveyance, 16 May 1849, recorded 19 February 1850, Deed Book C-12, pp. 
538-539, RMC). 

Smith then began to divide Curriboo Plantation, selling off 1300 acres to the 
North Eastern Rail Road Company in 1858 (Smith to North Eastern Rail Road Com- 
pany, Title, 1 May 1858, Deed Book Y-12, pp. 278-279, RMC). In 1862, Smith 
conveyed 930 acres to H. Panzerbeiter, "being a portion of the plantation 
called 'Curriboo'" sold by John Harleston "and others" to Charles Macbeth, 
who then sold it to Smith himself. This tract included the dwelling house, 
according to an extant plat, and indicated that the "Remainder of Curribboo" 
lay southwest of the Public Road/Santee River Road (Smith to Panzerheiter, 
Title, October 1862, Deed Book R-14, pp. 334-335, RMC) (Figure 21). Finally, 
in 1871, Smith sold 30 acres to Jacob V. Welch, described as "that Tract of 
Pine Land near St. Stephen Brick Church" (Smith to Welch, Title, 6 January 
1871, Deed Book A-16, p. 19, RMC). 

The section named "remainder of Curribboo" on the 1862 plat evidently re- 
mained separate from those tracts that subsequently assumed the name "Curri- 
boo." In 1897, Elizabeth R. Rickenbaker purchased three tracts of land, to- 
taling 1944 acres, which apparently constituted Curriboo Plantation. The 
deed read as follows: 

All that certain plantation or tract of land situate partly 
within and partly near the corporate limits of the Town of St. 
Stephens in the County of Berkeley in the said State 
containing Two thousand acres more or less on the southwestern 
side of the Santee River and bounded Northwest by lands 
formerly of Thurston known as "Yahan" South west by rest of 
"Curriboo" formerly of Chas. Macbeth and South East by lands 
late of Owen and others and consisting of three parcels 
according to a subdivision made for the purposes of the sale 
aforesaid. . .all of said plantation being called "Curriboo" 
(Court of Common Pleas, 21 January 1897, Deed Book Q-54, pp. 
53-54, RMC.BC). 



B-7 



Rickenbaker, in 1905, sold the three tracts to Amarentha Carter, explaining 
further that "said plantation was formerly of my late husband Augustus M. 
Rickenbaker dec'd; and in the settlement of his estate, was sold and conveyed 
to me by Joseph Stoppelbein, Esquire, Master in the Court of Common Pleas, 
for the County of Berkeley. . .January 21, 1897" (Rickenbaker to Carter, Ti- 
tle, 21 November 1905, Deed Book A-23, pp. 145-146, RMC.BC). 

In 1912, Carter sold 1100 acres "known as the 'Curriboo' Plantation" to W. H. 
Lorenz (Carter to Lorenz, Title, 20 February 1912, Deed Book A-37, p. 57, 
RMC.BC). In 1914, Lorenz sold D. M. Davis 745 acres "partly in the corporate 
limits of the Town of St. Stephens, known as Curriboo plantation formerly 
owned by Mrs. A. E. Carter" bounding "North by lands of Montague and Tucker 
and Piatt and Browder, East by lands of the Atlantic Coast Line R. R. Co., 
and of W. H. Ingram, South by the Public Road leading from St. Stephens to 
Pineville and West by Public Road leading from St. Stephens to Pineville and 
lands of Browder, Piatt and Montague and Tucker" (Lorenz to Davis, Title, 10 
August 1914, Deed Book A-41 , p. 192, RMC.BC). Davis sold the entire 745-acre 
tract to J. E. Bell of Berkeley County in 1918 (Davis to Bell, Title, 16 De- 
cember 1918, Deed Book A-47, p. 200, RMC.BC). Bell then sold the 745-acre 
parcel to R. S. Bell of Williamsburg County the following year (Bell to Bell, 
Title, 29 November 1919, Deed Book A-48, p. 212, RMC.BC). 

One month later, in December 1919, Bell mortgaged the 745 acres "being known 
as a portion of Curriboo Plantation," which he had acquired from J. E. Bell, 
to W. Y. Strong (Mortgage of Real Estate, 27 December 1919, Mortgages B-17, 
p. 260, RMC.BC). Bell defaulted payment, and Strong purchased the land at 
public auction (Court of Common Pleas, 16 December 1922, Deed Book C23A, pp. 
339-341, RMC.BC). In 1940, Strong sold Joseph Cooper "All that Piece. . .of 
land. . .being. . .in the County of Berkeley. . .containing seven hundred and 
forty-five (745) acres, more or less, being known as a portion of the Curri- 
boo Plantation and bounded on the North by lands of Santee Swamp; on the East 
by lands of Mrs. Ingram; on the South by road known as the Santee River Road, 
and on the West by lands of Piatt. ..." (Strong to Cooper, Title, 31 Octo- 
ber 1940, Deed Book A-66, p. 347, RMC.BC). 



APPENDIX C 

SOIL TESTS 

BIO-CHEM ANALYSTS, INC. 



TABLE C-l. Site 38BK75 



C-2 





Total Organic 


Total Organic 


Total Organic 




Carbon 


Nitrogen 


Phosphate 


75-F30-2 


0.8550% 


1.098% 


0.0064% 


75-F31-2 


0.3450% 


1.335% 


0.0041% 


75-F31-3 


0.2250% 


1.398% 


0.0001% 


75-F32 


0.2550% 


1.300% 


0.0023% 


75B-F2-6 


0.3600% 


1.403% 


0.0001% 


75B-F5-19 


0.3600% 


1.367% 


0.0020% 


75B-F5-21 


0.4500% 


1.491% 


0.0046% 


75B-F5-23 


0.4800% 


1.586% 


0.0001% 


75B-F5-25 


0.4050% 


1.450% 


0.0007% 


75B-F5-27 


0.4950% 


1.532% 


0.0028% 


75B-F22-2 


0.2700% 


1.452% 


0.0015% 


75B-F25-2 


0.3000% 


1.537% 


0.0050% 


75B-F25-3 


0.6000% 


1.470% 


0.0043% 


75B-F27-3 


0.2400% 


1.472% 


0.0050% 


75B-F29-5 


0.7500% 


1.487% 


0.5901% 


75B-F29-6 


0.4800% 


1.423% 


0.0048% 


75B-F29-7 


0.2550% 


1.484% 


0.0009% 


75F-31-1 (Matrix) 


0.2850% 


1.454% 


0.0001% 


75T-U1-5 Level 2 


0.6000% 


1.584% 


0.0002% 


75T-U1-5 Level 3 


0.2850% 


1.492% 


0.0040% 


75T-U5-5 Level 2 


0.4050% 


1.485% 


0.0038% 


75T-U5-5 Level 3 


0.2100% 


1.334% 


0.0001% 


75-F1-1 


0.1950% 


1.615% 


0.0065% 



TABLE C-2. Site 38BK76 



C-3 





Total Organic 


Total Organic 


Total Organic 




Carbon 


Nitrogen 


Phosphate 


76-F2-2 


1.0200% 


1.477% 


0.0517% 


76-F5-1 


0.3900% 


1.485% 


0.0001% 


76-F7 


0.3750% 


1.572% 


0.0020% 


76-F8-1 


0.9450% 


1.637% 


0.0878% 


76-F8-2 


0.7200% 


1.876% 


0.0960% 


76-F10-1 


0.2550% 


1.228% 


0.0600% 


76-F12-1 


0.3750% 


1.534% 


0.0297% 


76-F13 


0.3300% 


1.560% 


0.0276% 


76-F14-1 


0.3450% 


1.556% 


0.1752% 


76A-B9-3 


0.2100% 


1.500% 


0.0410% 


76A-F33 


0.7200% 


1.511% 


0.0268% 


76A-F39-1 


0.2850% 


1.399% 


0.1445% 


76A-U19-4 


0.3600% 


1.516% 


0.0177% 


76A-U26-1 


2.1900% 


1.475% 


0.0001% 


76A-U26-2 


0.5250% 


1.635% 


0.0262% 


76B-B20-1 


0.5550% 


1.457% 


0.0001% 


76B-F82 


0.8700% 


1.408% 


0.0001% 


76B-U39-2 


0.1650% 


1.403% 


0.0001% 


76D-F17-1 


0.2850% 


1.509% 


0.0354% 


76T-U9-3 


0.2400% 


1.553% 


0.0001% 



TABLE C-3. Site 38BK245 



C-4 





Total Organic 


Total Organic 


Total Organic 




Carbon 


Nitrogen 


Phosphate 


245-F3 


0.2700% 


1.414% 


0.0085% 


245-F7 


0.7800% 


1.502% 


0.0001% 


245-F12-3 


1.0050% 


1.474% 


0.0001% 


245-F12-4 


0.6750% 


1.563% 


0.0001% 


245-F62-1 


0.7500% 


1.580% 


0.0001% 


245-F63-1 


2.2500% 


1.552% 


0.0001% 


245A-AV (Inside Pot) 


0.3300% 


1.461% 


0.0001% 


245A-1-2 


0.4200% 


1.460% 


0.0040% 


245A-A55 


0.3600% 


1.471% 


0.0001% 


245B-F11-3 


0.3000% 


1.642% 


0.0001% 


245C-F3 


0.2700% 


1.527% 


0.0001% 


245C-F6 


0.3450% 


1.531% 


0.0001% 


245C-F7-1 


0.4500% 


1.497% 


0.0001% 


245C-F7-12 


4.3500% 


1.515% 


0.0001% 


245C-U3-1 (1) 


1.8500% 


1.639% 


0.0175% 


245C-U3-2 (1) 


1.4100% 


1.522% 


0.0036% 


245C-U3-3 (2) 


0.7800% 


1.477% 


0.0001% 



APPENDIX D 

ARCHITECTURAL HISTORIAN'S REPORT 

LANE GREENE 



D-2 



Have we not in Africa and in Spain walls of earth, known as 
"formocean" walls? From the fact that they are moulded, 
rather than built, by enclosing earth within a frame of 
boards, constructed on either side. These walls will last for 
centuries, are proof against rain, wind, and fire, and are 
superior in solidity to any cement. Even at this day Spai.. 
still holds watch-towers that were erected by Hannibal. 
Pliny's Natural History , Book XXV, Chapter xviii. 

The principle of building "walls of earth" is ancient, known to have been 
employed by the Romans before the time of Christ. The principle, broadly 
defined, is based on the fact that when certain suitable earth materials with 
the right moisture content are tightly compressed, they cohere to form a fair- 
ly hard, strong and solid body. The means by which the necessary cohesion is 
obtained may be either by external compaction of a relatively dry earth mix- 
ture contained within a system of formwork, or by the natural drying process 
of water from a wetter earth mixture. Earth walling techniques may thus be 
classified according to how this cohesion is obtained. 

The techniques can be classified into two major categories as follows: Pi se 
de terre or rammed earth, which involves the use of external compaction of 
the earth mixture within a system of slippable formwork, known as shuttering; 
and Cob or clay lump, which is the traditional method of building without 
shuttering. 

Pise de terre or rammed earth is a very old and simple method of building. 
The method consists of compacting the earth mixture by ramming while con- 
taining the mixture within the shuttering. The shuttering is raised, lift by 
lift, until the wall reaches its full height. This technique has been prac- 
ticed all over the world, and large structures can be found from the Rhone 
Valley in France, which have stood for 400 years, to the Church of the Holy 
Cross in Sumter, South Carolina, built between 1840 and 1850. 

Cob or clay lump walls differ from those of Pise de terre in the moisture 
content and plasticity of the material and in the absence of any shuttering. 
The wall is built up by the simple process of pitching on a soft but cohesive 
mixture of earth and straw, in layers, until the wall has reached its full 
height. Tapia and wattle and daub are but variations of Cob walling, as is 
the production of adobe or sun dried brick, although forming is used to con- 
tain the plastic mix in the latter. In all variations of Cob walling the 
underlying principle is that cohesion is achieved by the "silting" action on 
drying by evaporation — small particles settling in between larger particles 
to form a cohesive mass. 

Based on the archaeological evidence gathered at the Cooper River project 
sites it appears that a great many of the identifiable structures were 
originally constructed by the process of Cob walling. These structures have 
been identified as slave quarters and assuming (1) the correctness of this 
identification and (2) that the structures were built by their inhabitants, 
one must look to the origin and technological tradition of these builders. 



D-3 



This discussion assumes that the exact point of origin is of little conse- 
quence, since the ecological and cultural profile as it relates to building 
technology is very similar for a vast region of West Africa during the period 
of consideration. 

From Senegal all along the Guinea Coast and down the western 
coast of Central Africa there stretches a zone with remarkable 
consistencies. Along this more than 4,000 miles of coast line 
and a considerable distance inland the same rain forest en- 
vironment is encountered. The similarity of ecology is 
matched by the presence of root crop agriculture — and that 
musical systems are also consistent in this region. It is not 
surprising, then, that the architecture of this far-flung zone 
should possess basic similarities, too. The "rectangular, 
gable-roofed hut" is constantly identified with the rain 
forest environment. (Vlach 1978:124) 

Almost every variation of earth walling technique can be found in contempor- 
ary West Africa. In Zaire and Cameroon, walls are commonly built of Pise de 
terre and sun-dried bricks. In Angola, wattle and daub rather than Pise de 
terre is traditional; although all techniques, including sun-dried bricks are 
used in construction. In Nigeria the traditional use of earth for building 
walls is not confined to small houses only, for in the Haussa district, and 
especially in Kano, large buildings are built of native cone shaped sun-dried 
brick known as Tubal i which are laid with the large end of the cone facing 
alternately inward and outward and plastered with a clay covering applied to 
the wall as the work proceeds. 

Common to almost all cultures of West Central Africa, however, is the build- 
ing process whereby a lateritic earth, usually known as "swish", is kneaded 
into lumps about 6 to 8 inches 1n diameter, which are thrown down upon each 
other and pounded to form a compressed course about 12 inches wide by 12 
inches high. The wall being brought up, course upon course, in this fashion 
is then usually plastered, at least outside, with a variety of clay mixtures. 
Very often, and especially in the vaulted construction of the Haussa district 
of Nigeria, a form of reinforcement 1s provided by sticks, poles or palm 
fronds. Contemporary examples of almost identical building technology of 
walls built in this manner can be found in the rural areas from the Dogon 
region of Mali in the inland northwest to Angola in the coastal southwest. 

Local technological preferences do little to alter the underlying building 
technology from district to district within this larger region and as Clough 
Williams-Ellis stated in 1919; "The older and more skilled craftsmen in 
Nigeria do not travel far, and instruct only their own children, with the 
result that after a few generations the methods (of construction) differ only 
slightly from city to city" (Williams-Ellis et al . 1919:149). 



D-4 



Vlach and others have documented the contribution of African building con- 
cepts in the new world in the first half of the eighteenth century. It is 
reasonable that a people with limited, but uniform technology, transported 
from one semi-tropical environment to another with similar building materials 
would in fact build their own first structures in the style and with the 
technology with which they were familiar. 

There remain in this country a few examples of buildings known to have been 
built by slaves of West African origin which are closely related in techno- 
logy and appearance to certain styles of West African architecture. Note- 
worthy are three houses at Melrose Plantation, located on the Cane River near 
Natchitoches, Louisiana. Built in the late eighteenth century by Marie 
Therese Quan Quan, a former slave, two of the remaining structures, origi- 
nally called the Yucca and African houses, can be readily identified with 
rectangular houses having rammed earth walls and Bamil eke-type sloping roofs 
common to the regions of contemporary Zaire and Cameroon. 

Although built of brick, the round slave quarters at Keswick Plantation near 
Midlothian, Virginia, provides a yery fine example of the cylindrical or bee- 
hive houses common throughout sub-Saharan Africa and demonstrate the influ- 
ence of transplanted African architecture on colonial plantation buildings. 

The interpretation of the physical evidence of structures gathered at the 
Cooper River project site must be considered as conjectural in that little or 
nothing of the original building fabric was recovered. The evidence forms 
the basis for the hypothesis but alone is insufficient to support it. The 
following suppositions, by building elements, are based more on logic and the 
application of similar technological parameters than on actual facts. 

Foundations 

The system of trenches and postmolds recorded at Yaughan and Curriboo is not 
completely consistent with the implied technology. The postmolds within, 
along side and outside the line of trenches is open for further research but 
for the scope of this report will be considered as evolutionary within the 
life of the structure, i.e., the original configuration most probably having 
been a system of vertical wood post reinforcement within the thickness of the 
Cob walling. 

The function of the trenches is interpreted as a structural consideration to 
account for lateral thrust imposed on the side walls of the various struc- 
tures and in turn suggest a gabled roof configuration. The arrangement of 
the trenches in parallel pairs without end returns to complete the rectangle 
further supports this hypothesis. 

Walls 

The wall construction implied by the evidence is mud or Cob walling, most 
probably constructed as detailed earlier in this section of the report. 
Unfortunately, the cohesion of mud walling is a reversible process when 



D-5 



exposed to weather action and this has destroyed any measurable evidence. 
The assumption of mud wall construction with vertical wood post reinforcement 
is consistent with both the traditional technology of the builders and the 
available materials. The top of the side walls would most probably have 
included a wood log plate member to which roof rafters would have been se- 
cured. These plate members may have been secured by being embedded in the 
top of the last mud course, or by being lashed to the projecting vertical 
wood post reinforcing members, or both. 

The end walls of the structures are less easily understood. As stated ear- 
lier, the foundation can be interpreted as implying a gabled roof configura- 
tion. The construction of a gabled end wall in mud is, however, not only 
difficult but inconsistent with both the need to protect such walls by a roof 
overhang and the African tradition of hipped or partially hipped roof con- 
struction. The end walls may have been gabled mud walls, flat topped mud 
walls, walls of some other material such as reed matting or entirely open. 
The evidence is insufficient for conclusion. 

Roofs 

Possible roof configurations are discussed above. Probable roof construction 
would have employed wood pole rafters bearing on wall log plates with over- 
hanging eaves and a substantial pitch (possibly 12 on 12). Pole purlins 
would have then been secured to the rafters at some interval consistent with 
the roof covering and the roof covering applied. Possible roof coverings 
would include: thatched palmetto fronds, thatched sweetgrass, and split 
boards. Tradition and availability of materials would indicate palmetto 
fronds as the most probable choice and split boards as the least probable. 

Room Arangement/Building Proportions 

Comparative building proportions are treated statistically elsewhere in this 
report and generally fall within the established traditional African pref- 
erence for the two room, double square module. These "duplex" modules would 
have most likely housed two family groups - one family, one room. No evi- 
dence of indoor fireplaces or open hearths was discovered, leading to the 
assumption that cooking was done either out-of-doors or in some yet unident- 
ified central cook house or cook shed. 

Wall Openings 

Construction difficulty, tradition, and security all indicate that wall 
openings would have been minimal. A single doorway perhaps being the only 
source of entry, light, and ventilation. Window openings, if present, would 
most probably have been small and located at the top of the wall in order to 
eliminate the need for constructing a lintle over the opening. Doorways, 
similarly, would probably have been narrow but full height of the wall to 
avoid the problem of constructing the lintle over the opening. Doors and 
window shutters would probably have been later refinements to the original 
unsecured openings. Evidence was recovered to indicate that very little 
window glass was present in any of the structures. 



D-6 



Non-conforming Structures 

A certain number of building imprints did not conform to the preceding 
interpretation in that they did not contain trench foundation features. The 
possible superstructures for these structures could include: open sided 
sheds, framed and sheathed structures, and different technology of mud wall 
construction. A certain percentage of open sided, roofed sheds would be 
expected in any such agrarian compound. Construction of these sheds would 
most probably have been a simple post and lintle system with roof 
construction similar to that indicated earlier. 

Other structures indicate by rectangular or square postmolds that they were 
most probably much more technologically refined and were probably frame 
structures with wood plank sheathing. No speculation is offered as to roof 
construction or roof coverings for structures within this category. 

A possibility of mud wall structures constructed without trench foundations 
is also possible for structures within this category, in which case the 
preceding interpretations are applicable. 



D-7 



BIBLIOGRAPHY 

Merrill, Anthony F. 

1947 The Rammed-Earth House , Harper Brothers, Publishers, New York 

Ylach, John Michael 

1978 The Afro-American Tradition in Decorative Arts . The Cleveland 
Museum of Art, Cleveland. 

Williams-Ellis, Clough; John and Elizabeth Eastwick-Field 

1919 Building in Cob, Pise, and Stabilized Earth . Country Life 
Limited, London. 



APPENDIX E 
ARTIFACT APPENDIX 



E-2 



This discussion of the artifacts includes those types which were either 
minority types or types which by themselves provided little conclusive data 
on the slave occupation. Some of the data presented here has been used in 
the body of the report to compare with Colono ceramics or to illustrate the 
function of features and structures. The purpose of this appendix is pri- 
marily to give the background necessary to understand the data presented in 
the artifact patterns in Chapters VII and IX. 

Non-local Ceramics 

Porcelain was present at all sites and was divided into four types, Blue on 
White Oriental Porcelain, Polychrome Overglaze Oriental Porcelain, Plain 
Oriental Porcelain, and European Porcelain. A breakdown into more types, as 
done by South (1977a:210) was impossible due to the small sherd size. The 
size of the sherds prohibited any detailed study of motif, upon which much of 
South' s and Noel Hume's (1972 and 1978) studies appear to be based. Our 
typology more closely followed that of Miller and Stone (1970:82-94), which 
is more practical when dealing with small fragments. 

Sorting criteria for the four types were based on glaze and paste properties 
to differentiate Oriental from European types and decoration techniques to 
distinguish types within the larger groupings. Oriental porcelain was identi- 
fied by a hard, concoi dally fracturing paste with little or no distinct trans- 
lucent glaze on the interior and exterior surfaces. European porcelain, so- 
called because the specific country of origin is unknown, was distinguished 
on the basis of a softer paste and a distinct glaze. Only three sherds of 
European porcelain were found. 

The three Oriental porcelain types were differentiated by handpainted blue on 
white underglaze decoration, handpainted polychrome overglaze decoration, and 
plain Oriental sherds exhibiting no decoration. None of the European porce- 
lain sherds showed decoration. 

Stoneware was divided into 14 types based on paste color and surface treat- 
ment. Two of these 14 types were unidentified stoneware, one gray bodied and 
the other brown, which may have been late on the sites and may have represent- 
ed a variety of American stoneware types. The wide variety of surface finish 
and the overall small quantity of these- types made a detailed typological 
study impossible. The descriptions for British Brown, Nottingham, Burslem, 
Rhenish or Westerwald, Black Basalt, and all but one of the White Salt Glazed 
Stonewares are adequately described in the literature (South 1977, Noel Hume 
1978, Miller and Stone 1970), and do not require further description here. 
However, three types of Refined Red Stoneware and one type of White Salt 
Glazed Stoneware deserve special mention. 

The definition of unglazed Refined Red Stoneware followed that given by 
Miller and Stone (1970:70-81). Along with this type were are two other 
distinct types with the same paste: Clear Glazed Red Stoneware and Marbled 
Glaze Red Stoneware. The first was exactly like Refined Red Stoneware in 
paste, texture, and porosity, but occasionally had parallel wavy lines 



E-3 



incised into the body before application of the clear, probably lead, glaze. 
The second type had the same paste, texture, and porosity as Refined Red 
Stoneware, but had a glaze or more properly a slip glaze, which gave the 
surface an agateware quality. Both of these types were minority types. The 
first was represented by three sherds and the second by two. 

White Salt Glazed Stoneware was represented by the easily recognizable Plain 
White Salt Glazed plates, by Scratch Blue, Debased Scratch Blue, and by one 
example of a fourth type not described by South (1977), but noted by Miller 
and Stone (1970:72-74), i.e. Handpainted Polychrome White Salt Glazed Stone- 
ware. This single sherd had an overglaze floral motif. 

Earthenwares were divided into several subcategories based on paste color and 
texture, and surface finish. These were Refined Earthenware, Redware, 
Slipware, Creamware, Pearl ware, Whiteware, and Coarse Earthenware. 

Refined Earthenware included two types, Jackfield and Plain Agateware, which 
have been adequately described in the literature (South 1977:210). Coarse 
Earthenware was divided into two types already established in the literature: 
Buckley, a coarse Agateware with a laminated dark red and yellowish paste, 
and North Devon, a grit tempered type with a greenish glaze. Buckley was 
represented by four sherds and North Devon by three, and were definitely 
minority types. 

Creamware and Pearl ware accounted for 48 percent of the nonlocal ceramics at 
all three sites. Sorting criteria are fairly well established for the types 
included in these groups (South 1977, Noel Hume 1972 and 1978). Virtually 
all of the Plain Creamware, including "Queen's Ware, Royal Pattern", etc., 
found at the sites was light yellow; only two sherds of dark yellow were re- 
covered. No Carolina Creamware was recovered, and only one sherd of green 
Edged Creamware was found. The remaining Creamware types were classified on 
the basis of easily recognized and replicable criteria, which may or may not 
correspond to names used by various authors. For this reason, more complete 
descriptions of these types are included here. 

Clouded Creamware refers to creamware vessels decorated with brown slipglaze. 
Noel Hume (1972:125) refered to this as "clouded" or "tortoise shell" decora- 
tion (also Smith 1967). South (1977:211) also refered to this type as 
Clouded Ware, although Miller and Stone (1970:64) used the term 
"Whieldon-Wedgewood type". Polychrome Creamware (Miller and Stone 1970:48) 
was an overglaze polychrome, generally with yellow, black, purple, and brown 
hand painted decoration . The handpainted motifs were necessarily incomplete 
on the Cooper River material, but appeared to be lines and curves making up 
larger geometric and perhaps floral designs. South (1977:212) refered to 
this as "Overglaze Enamelled Handpainted Creamware". Marbled Creamware des- 
cribed the swirling together of various opaque colors (often dark and light 
brown and white slips over a base of light brown) 1n restricted areas of the 
exterior surface. This type appeared to be or develop into, a subset of the 
annular types since the decorated zones were often restricted by annular de- 
signs above and below. South (1977:212) used the term "Finger Painted Wares" 
to describe this type. Green Creamware was best described and illustrated by 



E-4 



Miller and Stone (1970:Figure 67) and in our sample the type had a green slip 
incorporating black dots in black circles under the glaze. One sherd of 
Green Creamware had a nonporous body which may have indicated overfiring or 
post-use firing. 

The Pearl ware types had many of the same type names as Creamware. This 
should not be surprising since Pearl ware was in many respects a continuation 
of Creamware. The sorting criteria for Edged Pearl ware with the addition of 
blue edges; Marbled, with the addition of reds and blues; and Polychrome were 
essentially the same as for the Creamware types. However, there were new 
types in Pearl ware not appearing in Creamware. Blue Handpainted Pearlware 
(South 1977:212) and a Brown Handpainted Pearlware were added in the collec- 
tion. Also for the first time in the assemblage Transfer Printed (usually 
blue) decoration appeared. Annular Pearlware with and without Marbling and 
Mocha decoration increased greatly over its rare use in Creamware. Eleven 
percent of the Pearlware sherds had annular decoration compared to 16 percent 
for Transfer Printed and 16 percent for Edged Pearlware. Plain annular dec- 
oration (excluding Marbled or Mocha) accounted for only 5 percent of the 
Pearlware sherds. This ran counter to the high frequency of annular types 
noted by Otto (1975:162) at slave quarters on St. Simons. Forty-seven per- 
cent of the total Pearlware sherds had no decoration and were catalogued as 
Plain Pearlware. These sherds were probably from decorated vessels for the 
most part (Miller 1980), but could not be identified as belonging to one type 
or another. 

A major problem encountered whenever Creamware, Pearlware, and Whiteware are 
found together is discriminating between the plain types of these wares 
(Price, 1979). As noted by Noel Hume (1972:217-254) and others, the yellow 
glaze and offwhite cast of the Creamware body evolved into a lighter yellow 
glaze and whiter body in some potteries by the time of the introduction of 
Pearlware. Pearlware was an attempt to make ceramics more reminiscent of 
porcelain and to achieve this effect cobalt was added to the glaze to reduce 
the yellow cast, and at the same time the whiter body (from the addition of 
chert to the paste (Noel Hume 1972:233)) was developed. The result was a 
whiter ceramic, often with a bluish cast where the glaze puddled. Both 
Creamware and Pearlware were here by the cast of the glaze (yellow to yel- 
lowish green for Creamware and robin's egg blue for Pearlware) and to a 
lesser extent by body color (a warm offwhite for Creamware and white for 
Pearlware). If it were not for whiteware, the problem of differentiation 
would have been a simple one. Whiteware presented additional problems, a 
variety of glaze tints, body colors, and the problem of porosity. 

It had been originally thought that porosity was exclusive or nearly exclu- 
sive to Creamware and Pearlware, and that, therefore, any sherds exhibiting 
porosity were not Whiteware by default. Although mid to late nineteenth (and 
all twentieth) century Whitewares tended to become more and more nonporous 
over time; in the period of the Pearlware to Whiteware transition (first half 
of the nineteenth century), porosity is not a useful tool for differentia- 
tion. This was forcefully brought home when complete sets of Whiteware from 
the Washington D.C. Civic Center (Garrow et al. 1981) site were compared 






E-5 



with the Cooper River material. These vessels, with makers marks identifying 
them as "Stone China", were as porous as known pieces of Creamware and 
Pearl ware; and, like Creamware and Pearl ware, exhibited a range of porosity. 
It was finally decided for purposes of classification that sherds exhibiting 
no porosity would be classified as Whiteware, while porous sherds would 
require other sorting criteria. 

The other criteria were body and glaze color. In some cases, the Whiteware 
body, though porous, was whiter than Pearl ware. There was a dead white 
quality about Whiteware difficult to describe. Unfortunately, this distinc- 
tion was too small to be objectively determined by the Munsell Color Charts 
(1975). The second criterion was glaze color, which was influenced by the 
underlying body color. Rather than the robin's egg blue of Pearl ware, 
Whiteware had a grayer cast and was not as "warm" as Pearl ware. This 
distinction, too, was impossible to detect with standard Munsell Color Charts 
(1975). It was felt that especially on earlier Whitewares (until mid- 
century), the glaze was identical or similar to that of Pearlware, but that 
the dead white body played a large role in changing the. quality of the blue 
cast to a grayer tint. Miller (1980) corroborates this impression and re- 
searched the question of the Pearl ware-Whiteware transition much more tho- 
roughly than is required here. Lofstrom (1976:8) felt, however, that the 
shift in the blue tinted glaze on Whiteware was the result of the development 
of a lead free glaze which did not come into popular use until the early 
1830s. We tend to agree with Miller's approach that the development of 
Whiteware involved a series of changes in the glaze and the body, making any 
distinctions between the two types difficult and basically a useless exercise 
after the first quarter of the nineteenth century. The term Whiteware re- 
ferred to the early Whiteware transition sherds which could no be clearly 
separated into Whiteware or Pearlware. True Whiteware and Ironstone were 
found in surface scatters at all sites and were not included in this 
analysis. 

The decorated types of transitional Whiteware were similar to those of 
Pearlware, as might be expected, although they were easier to sort from 
Pearlware than the Plain type. Blue Handpainted, Edged, Transfer Printed, 
and Polychrome Whiteware had essentially the same descriptions as those of 
the Pearlware types. 

Undoubtedly, many of the plain Whiteware sherds were identified as Pearlware 
as a result of the sorting criteria. The relative frequencies of the decor- 
ated varieties of Pearlware and Whiteware, which were easiest to consistently 
identify, were 95 percent and 5 percent respectively. However, among the 
plain types, the relative frequencies were 99.6 percent and .4 percent, 
respectively. This indicated that something over 4.6 percent of the plain 
Pearlware sherds were misidentlfied and should probably have been called 
Whiteware. Since the Whiteware sherds were coeval with the late Pearlware 
sherds, such mistakes in classification probably did not greatly affect 
chronological studies, and In fact may have enhanced them, since mean ceramic 
dates for Whiteware all postdated the terminal occupation of the sites (South 
1977, Bartovics 1977, and Lofstrom 1976). 

Tin glazed earthenware, referred to here as Delft, was classified in a manner 
similar to that used by Miller and Stone (1970). All sherds with a soft 



E-6 



porous body, a tin glaze, and no decoration were included in the Plain type 
with the exception of sherds definitely associated with Debased Rouen Faience 
as described by South (1977) and Miller and Stone (1970). Decorated sherds 
were sorted by the technique of decoration into types. Further division into 
varieties based on motif was impossible due to the small size of the sherds. 
All sherds with handpainted blue decoration were included in the Blue and 
White type. Those with two or more different colors, usually a combination 
ofblue, red, or green, were included in the Polychrome type. Eleven sherds 
had powdered manganese purple on the interior or the interior and exterior. 
Rims of this Powdered Delft often had a poorly painted horizontal blue line 
on the interior as described by Miller and Stone (1970:40-42). Only two 
sherds definitely attributable to Debased Rouen Faience were recovered. 
These had a thin poorly applied tin glaze allowing the pink body to show 
through, giving an uneven pinkish cast to the glaze. The interior rims were 
decorated with wavy yellow lines. As various authors have stated, without 
complete vessels, attributions as to country of manufacture are generally 
misleading. Unfortunately, we have come to the same conclusion here. 

The remaining ceramics and other Kkitchen Group artifacts are described in 
Chapter VIII as they provide more direct evidence on slave lifeways and 
culture change. 

Architecture Group 

The architecture group artifacts include nails, bricks, mortar, daub, flat 
glass, and hardware. Flat glass was examined to detect crown, cylinder, and 
plate glass. However, most sherds were so small that such a distinction 
could not be made. For this reason, all flat glass was lumped together. 
Such glass, presumably used for windows, never accounted for more than 3 to 8 
percent of the architecture group artifacts, except at 245C, where it made up 
over 17 percent. While no attempt was made to reconstruct window panes, it 
can be safely stated that none of the structures except 245C had enough glass 
to make more than one pane. Due to the natural variation in crown glass, a 
minimum pane count was not attempted either. Poor preservation at the sites 
also precluded identification of more than a small percentage of the nails. 
Although these were examined and, in some cases x-rayed, the nails could only 
be broken down into wrought, cut, and unidentifiable. 

The door lock plate at 38BK75 1s virtually identical to a plate stock-lock of 
the eighteenth century illustrated in Figure E-l and by Noel Hume 
(1978:Figure 77b). Unfortunately, it was not found in association with a 
structure. Other architectural hardware consisted of hinges, pintles, a 
possible shutter dog at Structure 245C, and door locks (Figures E-l and E-2). 

Wood artifacts included posts, dfscussed by structure 1n Chapter VII, and a 
wooden peg, measuring 2" by 1/4" recovered from the cellar at Structure 245C. 
While wood definitely played an important role in the architecture, preserva- 
tion prevented analysis of woodworking methods to any extent. 

The bricks are discussed in some detail since they were a source of income at 
Curriboo Plantation. All of the bricks at Curriboo and Yaughan were appar- 
ently made in the same way. A frame, probably of wood, was constructed that 
was open on the top and bottom. Sand was sprinkled on a flat surface, 



E-7 





CENTIMETERS 



OOOR HINGE 



CLOTHING IRON 




BRICK 



STOCK LOCK 





INCHES 



CENTIMETERS 



OOOR HINGES 



LEFT- SHUTTER DOG (?) 

RIGHT "DECORATIVE BRASS FIGURE E-l 

Architecture and 
Other Artifacts 



HORSE TACKLE 








INCHES 



CENTIM£TE«8 



SHUTTER HINGE, LOCK BOLT 



FURNITURE HANDLE, CLASP KNIFE 




CENTtl#ETER8 



STAMPED FLAT BRASS, BOTTLE SEAL, HINGE 
AND UMBRELLA STRUTS 



GOLD BUTTON 




I N C M If S 



Ce«TtMFTEI»8 





(NCHES 



CENTIMETERS 



CHEST LOCK 



KEY , JEW'S HARP 



FIGURE E-2 

Architecture, Kitchen, 
Furniture, Clothing and 
Personal Group Artifacts 



E-9 



CNJ 






m a; "O 

i en oj 

lu c •«- >> 

q: -m 

= j- c x 

u o 41 

>— i O T3 >i 

b_ Q M _Q 




E-10 




*a- .*: -o 
i u <v 
UO'p | 

_i «*- id 

UJ -t— es 

Ci ^i _ 

s u e x 

tr o a> 

u_ 00 •— _o 



a> 

M 

55 

"3 

o 

< 
to 



E-ll 



probably the bare ground, and the frame placed on the sand. Clay was mixed 
with water to a mortar-like consistency and then placed in the frame. Using 
his hand or a flexible tool, the brick maker wiped off the excess clay from 
the top of the frame, leaving it slightly indented and striated where sand 
and pebbles were dragged along the surface. The frame was removed while the 
clay was still fairly soft, allowing the brick to slump and making it slight- 
ly wider at the base and slightly shorter. The original underneath side had 
impregnated sand from the sand placed on the ground before the frame was set 
up. Once sun or air dried, the shrunken bricks were collected and stacked 
for firing. 

Given the above (and it was clear that the bricks were not made in a closed 
mold, i.e. with an enclosed bottom), then the validity of the height, width, 
and length measurements should not have been equal. The variation in height 
was determined by two independent factors, how deeply a brick was gouged on 
the top and by subsequent slumping; and the width and length by only one 
variable, the amount of slumping. Therefore, the width and length, where the 
latter could be measured, were the most valid measurements for comparison. 

The data on the bricks was organized by width and height and when possible 
length. Comparable data on bricks could not be developed from coastal South 
Carolina, and information provided in Noel Hume (1978:81), refered to England 
and the mid-Atlantic region, i.e. Virginia. 

Data from the surface collection at the owner's house west of Site 38BK75 
(i.e. 388K75 Locus B) is given in Table E-l for comparison, but the sample 
was too small for statistical analysis. 



Table E-l. Brick Measurements (in inches) 





Structure 24 


5K 


Structure 245C 


75 Locus B 
(Owner's House) 






WIDTH 






Mean width 


4.11 




4.19 


4.29 


Standard deviation 


0.1224 




0.1956 


0.2103 


Number Measured 


131 


HEIGHT 


39 


8 


Mean height 


2.96 




2.96 


3.35 


Standard deviation 


0.2023 




0.2062 


0.0768 


Number Measured 


131 




39 


4 



E-12 



Mean length 


8.48 


8.39 


8.52 


Standard deviation 


0.2805 


0.2419 


N/A 


Number Measured 


3 


23 


1 



The conclusion drawn from these data was that the bricks produced at the kiln 
(Structure 245K) could have been used in the construction of the office 
(Structure 245C). 

Furniture Group 

This, along with the personal group, had the lowest frequency of any of the 
groups at the sites. This was undoubtedly due to the low incidence of slaves 
owning extensive amounts of material goods at Curriboo and Yaughan. 

The furniture hardware was primarily made up of tacks with a restricted as- 
sortment of small drawer or cabinet door pulls, latches, a hinge, and brass 
decorated items (Figures E-2 and E-5). The following chart gives a breakdown 
by site and type. 

Table E-2. Furniture Hardware 

Tacks Brass 

Brass. Brass Brass Decorative 
Brass Iron Other Hinge Pulls Latches Items Total 



38BK75 






2 


1 






2 


5 


38BK76 


2 


4 


2 




2 


1 


1 


12 


38BK245 


1 




2 






1 




4 



The "other" category of tacks were heavily oxidized and may have been iron. 
The tacks were included in furniture since all had large heads reminiscent of 
upholstery or decorative tacks. The latch parts consisted of a piece of 
stamped flat brass with a hook on the end, measuring approximately one inch 
and a loop with a screw shank into which a hook would have fit. The presumed 
drawer or door pulls were only fragments and may have been some other decora- 
tive item. The decorative items at 38BK75 were flat brass with stamped 
floral designs such as might have appeared on the corners of chests. At 
38BK76, there was a concave piece of brass with stamped decoration which may 
have attached to the end of a leg of furniture (or possibly a cane). Most of 
the material from Site 38BK76 came from Structure 76B. The remainder of the 
material was too infrequent at any of the sites to recognize any clear dis- 
tribution pattern. It should be noted that a chest lock is discussed in the 
section on the Activities Group locks following South (1977a). 



E-13 







#1 





A 



y / '■ 







%»~' A 



y&9 



INCHES 



CENTIMETERS 



TOP ROW -THIMBLE, DECORATIVE BRASS CANE TIP(?), 2 WHITE METAL BUTTONS 

MIODLE ROW -LEAO BUTTON, BRASS BUTTON, LEAO BUTTON, BRASS BUTTON, LEAD BUTTON 

BOTTOM ROW -2 WHITE METAL BUTTONS, LEAD BUTTON, 2 WHITE METAL BUTTONS 





CSNTIMSTENS 



LEFT TO RIGHT - GRAPHITE OBJECT, ASSORTED BUCKLES 



FIGURE E-5 
Furniture and 
Clothing Artifacts 



E-14 



Clothing Artifacts 

Clothing artifacts consisted primarily of buttons, although buckles, a thim- 
ble, straight pins, bale seals, and glass beads were also recovered (Figure 
E-5). The following table lists the buttons following the types proposed by 
South (1977:100) and one type from 01 sen (1963: Figure 1). 



Table E-3. Buttons 



South Type 



38BK75 



38BK76 



38BK245 



4 
6 
7 
8 
9 
11 
13 
18 
F* 
7 or 8 
Unknown 



Total 



"29" 



18" 



TT 



*01 sen's (1963) type F 

These button types were given general date ranges by Noel Hume (1978:88-92) 
and 01 sen (op. cit.). Types 4, 6, 7, 8, 9, and 13 dated between 1726 and 
1776. Type 11 dated between 1726 and 1865; Type 18 between 1800 and 1865; 
and Type F between 1812 and 1830. These dates are only approximate and, 
except for 01 sen's Type F, were dated from archaeological contexts. Never- 
theless, the following chart indicates that 38BK76 and 38BK245 were probably 
earlier than 38BK75 on the basis of the buttons. 



Date Range 



Table E-4. 
38BK75 



Dated Buttons 



Buttons 
38BK76 



38BK245 



1726-1776 
1726-1865 
1800-1865 
1812-1830 



20 

1 
3 

1 



25 

4 

1 



The preponderance of plain round buttons was not considered exceptional on 
slave sites. The gold button found at Structure 76A, however, was extraordi- 
nary. According to Lucille Weingarten (personal communication, 1979), the 



E-15 



button, "is of European origin, probably French, from a man's vest or waist- 
coat, pre-19th century". Isaac Cordes' inventory listed gold buttons at his 
death in 1745. From that time on, however, small personal items were not 
mentioned in the inventories. It is possible that the gold button from 
Structure 76A had originally been in the Cordes family. Other clothing group 
artifacts were not as numerous. 



Table E-5. Other Clothing Artifacts 
38BK75 38BK76 



38BK245 



Buckles 
Straight pins 
Thimble 
Bale seals 
Glass beads 



6 

1 
1 

20 



The straight pins were brass with applied beads. The thimble was notable for 
its resemblance to the "thimble impressed" decoration on a Colono sherd. The 
bale seals, one with brass wire attached, implied the processing and shipment 
of cotton. The beads are described separately in Appendix F, as they were 
analyzed by Marvin Smith. 

Personal Group 

With the furniture group, this group of artifacts had the lowest frequency of 
occurrence at any of the sites (Figure E-2). One coin was found for the en- 
tire project at Site 38BK75. This was a 2 reale piece minted in Mexico in 
either 1758 or 1768 and in very worn condition. Three keys were found at 
38BK76 and one at 38BK245. Other personal items were two umbrella struts 
from Feature F29 at 38BK75 and a brass finger ring from general excavation. 
At 38BK76B, there was a piece of twisted lead possibly used as a pencil and, 
on the surface, a piece of metal backed mirror. These artifacts gave mute 
testimony to the lack of material goods owned by the slaves at all three 
sites. 



APPENDIX F 

BEAD ANALYSIS REPORT 

MARVIN SMITH 



F-2 



GLASS BEADS FROM THE COOPER RIVER REDIVERSION CANAL PROJECT 



This report will present a descriptive and comparative analysis of 25 glass 
beads recovered from three plantation sites in the South Carolina Coastal 
Plain. Twenty-one of the beads came from one site, 38BK76, while two other 
sites, 38BK75 and 38BK245, accounted for the remaining four beads. Beads 
were recovered from the surface, from screened excavation units, and from 
features processed by flotation techniques. 

While this collection of beads is quite small, it is of importance since it 
is one of the few reported assemblages of eighteenth century beads from a 
Black plantation slave context. Generally, the beads are fairly typical of 
those traded to North American Indians. Following a brief discussion of bead 
manufacturing techniques and bead typology, the beads will be classified ac- 
cording to a descriptive typology. The beads will then be compared to those 
found on Indian sites in North America and a slave cemetery in Barbados. 

Manufacture of Glass Beads 

Two major manufacturing processes were in use during the eighteenth century: 
the hollow cane technique and the mandrel wound technique. In the hollow 
cane technique, a large bubble of glass is drawn out into a long tube, or 
"cane", which is then cut into short sections for beads. Frequently, such 
cane beads were then tumbled over heat with a polishing agent to round and 
smooth the beads, and are, therefore, known as tumbled cane beads. Mandrel 
wound beads are produced by winding a molten thread of glass around a spin- 
ning rod, or mandrel, until a suitable sized bead is built up. These beads 
can be further modified by pressing facets on the beads while they are still 
hot and plastic. See Good (1972) and Kidd (1979) for further discussion of 
bead manufacturing techniques. 

Bead Typology 

After beads are classified according to their manufacturing technique, they 
are further classified according to their structure. Simple beads are com- 
posed of one layer of glass, compound beads are composed of two or more 
layers of glass, and complex beads have applique or inset decorative ele- 
ments. Beads which are both compound and complex, that is those beads which 
are composed of two or more concentric layers of glass with inset decorative 
elements, are classified as composite (Stone 1974). 

Finally, beads are classified according to their colors. Since a standard 
color chart was not available to this author, color descriptions are general. 
Specific proveniences of all beads are listed in Table F-l. 

Bead Types 

Drawn Cane Beads 

Type 1. Tubular translucent blue untumbled cane bead of simple construction. 
One complete and one fragmentary specimen. Diameter: 7-8mm, Length: 24mm 
(Figure F-l). 



F-3 



Type 2. Tubular translucent green tumbled cane seed bead of simple construc- 
tion. Three specimens. Diameter: 3mm, Length: 2mm (Figure F-l). 

Type 3. Tubular translucent navy blue tumbled cane necklace bead of simple 
construction. One specimen. Diameter: 7.5 mm, Length: 6mm (Figure F-l). 

Type 4. Barrel shaped opaque turquoise blue tumbled cane necklace bead of 
simple construction. One specimen. Diameter: 8 mm, Length: 10.5 mm 
(Figure F-l). 

Type 5. Barrel shaped brick red tumbled cane necklace bead of compound con- 
struction. The bead consists of three layers: a thin clear layer, a thin 
brick red layer, and a translucent green core. Two specimens. Commonly 
referred to as a Comal ine d'Aleppo. Diameter: 7-8.5 mm, Length: 8.5-9 mm 
(Figure F-l). 

Type 6. Tubular tumbled Comal ine d'Aleppo. Shape variant of Type 5. One 
specimen. Diameter: 5 mm, Length: 14.5 mm (Figure F-l). 

Type 7. Comal ine d'Aleppo donut shaped seed bead. Size variant of Type 5. 
One specimen. Diameter: 3 mm, Length: 2 mm (Figure F-l). 

Type 8. Opaque white tumbled "pony" size cane bead of compound construction. 
The white bead has a clear glass overlay to add gloss. One specimen. 
Diameter: 4 mm, Length: 2.5 mm (Figure F-l). 

Type 9. Striped cane bead of untumbled composite construction. An off white 
core layer is covered by a thin layer of opaque white which has six brick red 
stripes made up of minute multiple canes which show as individual canes in 
some areas. The whole bead is covered with clear glass for gloss. One speci- 
men. Diameter: 6.5 mm, Length: 20 mm (Figure F-l). 

Wire Wound Beads 

Type 10. Clear barrel shaped wire wound bead of simple construction. Two 
specimens. Diameter: 11 mm, Length: 8-8.5 mm (Figure F-l). 

Type 11. Clear donut shaped wire wound bead of simple construction. One 
specimen. Diameter: 11 mm, Length: 6 mm (Figure F-l). 

Type 12. Opaque white olive shaped wire wound bead of simple construction. 
Often called a "barley corn bead". One specimen. Diameter: 4 mm, Length: 
7 mm (Figure F-l). 

Type 13. Fragment of a large translucent blue wire wound bead of simple 
construction. This bead was probably originally olive shaped. One frag- 
mentary specimen. No size recorded. (Figure F-l). 

Type 14. Translucent blue "drop" or barrel shaped wire wound bead of simple 
construction. One complete and two fragmentary specimens. Diameter: 9.5 
mm, Length: 10 mm (Figure F-l). 



F-4 



Type 15. Opaque turquoise blue spherical wire wound bead of simple con- 
struction. One specimen. Diameter: 4 mm, Length: 4 mm (Figure F-l). 

Type 16. Translucent blue tubular wire wound bead of simple construction. 
The bead has been marvered into a pentagonal cross section while plastic. 
One specimen. Diameter: 11 mm, Length: 11 mm (Figure F-l). 

Type 17. Translucent blue bead of simple, wire wound construction, with 
eight pressed facets and two unmodified ends. Commonly called a decahydral 
bead. One fragmentary specimen. Diameter: 10 mm, Length: 8 mm (Figure 
F-l). 

Type 18. Opaque Turquoise blue "funnel -shaped" wire wound bead of simple 
construction. One fragmentary specimen. Diameter: 4 mm, Length 6+ mm 
(Figure F-l). 

Comparative Analysis 

Table II presents a comparative analysis of the Cooper River project beads. 
Comparisons are made with the Newton Cemetery, a slave cemetery in Barbados 
ca. 1660-1775 (Handler and Lange 1978); the Guebert site, a Kaskaskia Indian 
site in Illinois 1719-1833 (Good 1972); Fort Moore, South Carolina, 1680-1763 
(Storey n.d.; Polhemus 1971); and a sequence of trade beads established from 
Wichita Indian sites in Texas for the period 1700-1850 (Harris and Harris 
1967). Many other Indian sites from the eighteenth century could have been 
utilized for comparative purposes, but the ones chosen cover a wide geograph- 
ical range and have well described samples of beads. 

It is clear that the assemblage of beads from the Cooper River Plantation 
sites dates to the eighteenth century. Virtually all of the beads have been 
found on eighteenth century Indian sites. Furthermore, the faceted beads 
typical of nineteenth century Black slave and Indian sites (Ascher and 
Fairbanks 1971; Fairbanks 1974; Good 1976) were not found. Most of the beads 
could easily be attributed to the early to mid-eighteenth century, although 
Harris and Harris (1967) indicate that virtually all of the types were common 
well into the nineteenth century. Only Type 12, the barleycorn bead, appears 
to date from the late eighteenth century; after 1767 according to Harris and 
Harris (1967). Interestingly enough, this bead was found inside the hole of 
a Type 10 bead. 

Type 15 and Type 18 may well date to the nineteenth century. A red counter- 
part to Type 18 has been found on Creek Indian sites (post 1836) in Oklahoma 
(Mary Elizabeth Good, personal communication). The wire wound bead Type 11 
has also been found in nineteenth century slave contexts at the Hermitage 
(Good 1976:Type R) . Nonetheless, the assemblage as a whole is most typical 
of the mid-eighteenth century. Mary Elizabeth Good was kind enough to study 
a slide of the beads, and she concurs that it is a mid-eighteenth century 
assemblage (Good, personal communication). 



F-5 



One tentative observation can be proposed. Certainly in the Cooper River 
project, seed beads were quite rare, although features were processed by fine 
flotation recovery techniques. Seed beads were also rare at the Hermitage 
(Good 1976) and were not recovered from the Rayfield Plantation (Ascher and 
Fairbanks 1971), although these latter sites both date from the nineteenth 
century. Apparently southern plantation slaves did not have time for bead 
embroidery, or they simply did not have access to the small beads. Just as 
clearly, bead necklaces were an important item of personal adornment. Un- 
fortunately, little is known of the slave's means of access to the beads. It 
is doubtful that they were bartered from Indians, since many of the sites 
known archaeologically were in areas largely depopulated by Native Americans. 
Perhaps beads were distributed as bonuses to hard working slaves. A thorough 
search of relevant historical materials might shed light on this problem. 

In conclusion, the collection of beads from the Cooper River project is 
important since it is the only sample of eighteenth century beads from a 
southern Black slave context. Larger samples of beads would be desirable, 
but it appears that bead necklaces were an important article of personal 
adornment, while articles of clothing embroidered with seed beads were not. 



F-6 



Table F-l. Provenience of Beads 

Type Catalogue Number 

1. 76B-U31-1; 245-L17-5 

2. 76D-F18; 761-0-5; 76-F8-1 

3. 76-F8-1 

4. 76-0-0 

5. 76-L10-5; 76B-U30-1 

6. 76-L10-0 

7. 76B-F102 

8. 76B-F88-3 

9. 76-L10-0 

10. 76-F2-4; 76A-U18-2 

11. 76-0-0 

12. 76-F2-4 

13. 76A-U22-1 

14. 76-L9-5; 76B-F102; 76B-U30-1 

15. 245-1-0 

16. 76B-U27-1 

17. 75B-F29-5 

18. 245-F62-1 



F-7 



Table F-2. Comparative Analysis Table 



Cooper 










River 








Wichita 


Project 




Fort Moore, 


Newton 


Sequence 


Type No. 


Guebert 


w • O • 


Cemetery 


#/Date 



1 

2 

3 
4 
5 
6 
7 
8 
9 

10 
11 
12 
13 
14 
15 
16 

17 
18 



122 


22 or 199 


59 


78 or 219 


90a 


77 


127 


9 


123 


8 variant 


127a 


8 


107a 


6 


— 


193 


49 


Probable 


43 


162 


39 




— 


244 or 360 


46 


? 


1 


occurs in 
white 



26v 





164/1740-1820 




10/1700-1836 




99/1740-1836 




57/1740-1820 


26y 


51/1700-1836 




45/1700-1836 


26o 


— 


26q 


93/No Date 




101/1767- 



66 



26k 



41/1700-1820 



F-8 



References Cited 



Ascher, Robert and Charles H. Fairbanks 

1971 Excavation of a Slave Cabin: Georgia, U.S.A. Historical 
Archaeology 5:3-17. 

Fairbanks, Charles H. 

1974 The Kingsley Slave Cabin in Duvall County, Florida, 1968. The 
Conference on Historic Site Archaeology Papers 7, 1972. 

Good, Mary Elizabeth 

1972 Guebert Site: An 18th Century, Historic Kaskaskia Indian Village 
in Randolph County, Illinois. Central states Archaeological So- 
cieties Memoi r II. 

1976 Glass Beads from the First Hermitage. In "An Archaeological and 
Historical Assessment of the First Hermitage", edited by Samuel 
D. Smith. Tennessee Department of Conservation, Division of 
Archaeology, Research Series 2:237-248. 

Handler, Jerome and Frederick W. Lange 

1978 Plantation Slavery in Barbados: An Archaeological and Historical 
Investigation . Harvard University Press, Cambridge. 

Harris, R. King, and Inus M. Harris 

1967 Trade Beads, Projectile Points, and Knives. IN A Pilot Study of 
Wichita Indian Archaeology and Ethnohi story edited by Robert E. 
Bell and others: pp. 129-158. Manuscript report, Southern 
Methodist University. 

Kidd, Kenneth E. 

1979 Glass Bead-Making from the Middle Ages to the Early 19th Century. 
History and Archaeology 30, Parks Canada, Ottawa. 

Polhemus, Richard 

1971 Excavations at Fort Moore-Savano Town (38AK4 & 5). The Institute 
of Archaeology and Anthropology Notebook II I (6): 132-133. 
University of South Carolina, Columbia. 

Stone, Lyle M. 

1974 Fort Michilimackinac, 1715-1781: An Archaeological Perspective 
on the Revolutionary Frontier. Publications of the Museum, 
Michigan State University, Anthropological Series 2, Lansing. 

Storey, C. E. 

n.d. Trade Beads and other Relics of Historical Interest Found at Fort 
Moore. Manuscript on file at the Augusta-Richmond County Museum, 
Augusta, Georgia. 



F-9 




an <s) 

=> -a 

U_ CO 



fO 






(V 






CM 0> 




00 


• m 


CO 


•■ 


— 00 


"~„ 


1*- 


YPES 
6,7, 


<M 


<£ 


T"> 


O 


£ 




""■ 




~o 


- 


^ 


f- 


o 




X . 






o W 




<n 




CO 

UJ 


UJ 


P £ 


a. 


> 


K H- 


> 


h- 


h- • 


i- 




U- «_ 




* 


UJ $ 


i 


5 


-1 O 


$ 


O 


*o 


o 


oe 


0C 


2 


<r 2 


o 


O 


" o 


oe 


H 


a. (j 




K 


O UJ 


X 


O 


H C/i 


H 


CD 



APPENDIX G 
ETHNOBOTANICAl REPORT 
PAUL GARDNER 



THE ANALYSIS AND INTERPRETATION OF PLANT REMAINS 
FROM THE YAUGHAN AND CURRIBOO PLANTATIONS, 
BERKELEY COUNTY, SOUTH CAROLINA 



by 

Paul S. Gardner 



Department of Anthropology 
and 
Research Laboratories of Anthropology 
University of North Carolina at Chapel Hill 



September, 1980 



6-3 



In 1979, in order to mitigate the destruction of historic antebellum plan- 
tations by the construction of a canal between the Santee and Cooper Rivers 
near Saint Stephens, South Carolina, archaeological excavations were under- 
taken by Soil Systems Incorporated under the auspices of the Interagency 
Archeological Service. A description of the overall excavation procedures 
used at the Yaughan Plantation (38BK75 and 38BK76) and at the Curriboo 
Plantation (38BK245) have been presented elsewhere and need not be repeated 
here. What is of concern to this study is the fact that the excavators made 
use of soil flotation to obtain plant remains preserved in the archaeological 
deposits. While the use of flotation procedures is becoming increasingly 
common on excavations of prehistoric archaeological sites, they have been 
slow to be accepted by historic sites archaeologists. This is unfortunate, 
as soil flotation can be productively carried out using a minimum of simple 
and inexpensive equipment. 

While mechanical froth flotation devices are best for dealing with large 
quantities of soil, much simpler equipment can be used on most sites with 
fruitful results. The simplest arrangement, known as the "bucket method", 
involves half filling a bucket with water, slowly pouring in a measured 
amount of soil, and agitating gently to free the charcoal which then is 
carried to the surface of the water. Then the water is carefully poured 
through a screen which catches the charcoal particles. Pouring should cease 
before the heavy residue in the bottom of the bucket enters the screen. Ad- 
ditional water can then be added to the bucket and the procedure repeated, 
until no further charcoal is freed from the soil. At this point, the remain- 
ing heavy residue can be water screened to retrieve other small remains. 

Another simple arrangement is known as the "immersion method". Here a mea- 
sured amount of soil is added to a bucket whose bottom has been replaced with 
a mesh screen. The bucket is half immersed into a larger tub of water (or a 
stream, river, etc.) and agitated. The charcoal which floats to the surface 
of the bucket is skimmed off with a tea strainer. A variant of this method 
was used at the Yaughan and Curriboo Plantations, but here a 55 gallon drum 
was halved and window screen stretched across it. Water was then added to 
the drum to cover the screen, and soil was then poured directly onto the 
screen and stirred. The charcoal thus freed was then skimmed off with a tea 
strainer. 

Regardless of which flotation techniques are used, a few simple guidelines 
should be adhered to in order to simplify analysis and to insure comparabi- 
lity of results from one site to the next. First, the charcoal retrieved by 
flotation should be allowed to dry slowly in order to mitigate breakage. 
This can best be accomplished by emptying the charcoal onto several thick- 
nesses of absorbent paper towels (a spray bottle of water works well in 
removing any adherent charcoal from the screen), and then enclosing this in a 
sheet of newspaper. This package can be labeled with waterproof ink and then 
placed outside of direct sunlight to dry. Second, the mesh size of the 
screen used in flotation should be recorded. Window screen has a mesh size 
of about 1.3 millimeters, and tea strainers have a mesh size of about 0.7 
millimeters to 1.0 millimeters. Better than either of these, however, are 
geologic sieve screens, which are more durable and have a more uniform mesh. 



G-4 



A mesh size of 0.5 millimeter to 0.7 millimeter is most suitable. Finally, 
the volume of each soil sample floated should be recorded. This allows the 
computation of standardized measures of the abundance of plant remains (grams 
of remains/liter and seeds/liter) which can be used for intersite compari- 
sons. * It is convenient to use a standardized measure — two to four liters 
generally works well with the non-mechanical devices, but the ideal volume 
will vary with soil conditions. (For a more detailed discussion of flotation 
devices and techniques, see Watson, 1976). 

Analytic Procedures 

The ethnobotanical analysis followed the standard procedures of the Research 
Laboratories of Anthropology, developed by Richard A. Yarnell (cf Yarnell 
1974). Briefly, these procedures are as follows. Each sample to be analyzed 
is weighed, then fractioned through a series of stacked geologic screens. 
This screening produces a set of subsamples, each composed of approximately 
equal sized particles which are more easily examined than the unsorted ma- 
terial. Each subsample is weighed, then examined under a variable power (7x 
to 30x) dissecting microscrope. All seeds are removed from each subsample 
and identified, to the most limited taxonomic level possible, usually genus. 
As the weight of the seeds is usually quite small, all seeds from a sample 
are combined and their aggregate weight reported. In addition, the counts of 
each seed type are reported. 

Identification of other plant remains is carried out for those remains great- 
er than 2.0 millimeters in size, and the weight of each category of material 
is taken. Unfortunately, remains smaller than 2.0 millimeters in size cannot 
be confidently identified. However, in order that the quantities of material 
reported might more accurately reflect the composition of the sample as a 
whole, the weights of the remains larger than 2.0 millimeters are extrapo- 
lated to the remains that are between 2.0 millimeters and 0.7 millimeter in 
size. This extrapolation cannot be extended to the material smaller than 0.7 
millimeter since this material is primarily fine dirt particles and rootlets 
with a disporportionately small amount of carbonized plant remains; therefore 
the weight of the non-seed material passing through the 0.7 millimeter screen 
is ignored and not entered into the tables. 

Flotation of the Yaughan and Curriboo Plantation feature fill was so success- 
ful in obtaining plant remains that it was not feasible, for reasons of time 
and budget, to examine every sample — the quantity of material was simply 
too great. It was therefore necessary to reduce the amount of material to be 
analyzed while still maintaining a data base sufficient to allow inferences 
to be drawn to the site as a whole. This was accomplished in the following 
manner. First, a preliminary examination of the samples showed that the post- 
hole and postmold samples contained little information. These samples were 
usually quite small — the fifty-two such samples from Curriboo Plantation 
had a median weight of only 0.96 gram — and were almost without exception 



1-These data were not recorded for the Yaughan and Curriboo Plantations; 
therefore, these statistics have not been calculated in this report. 



6-5 



quite trashy, containing more rootlets and dirt than carbonized material. 
Hence the postmold and posthole material was not analyzed except for four 
samples from the Yaughan Plantation. Likewise, the samples from other 
non-pit features were slighted for the same reasons. On the other hand, the 
samples from trash pits were generally large and, compared to other samples, 
relatively clean of non-carbonized remains. In addition they are the type of 
feature most likely to contain carbonized plant remains. For these reasons 
it was decided to focus the analysis on this class of feature. Unfortunate- 
ly, it was not possible to analyze even the entirety of the feature material. 
Instead, for the particularly large features, the samples from one-half or 
one-fourth of the feature would be analyzed. In this fashion it was possible 
to analyze at least a portion of the material from each trash pit from the 
three sites. While a greater quantity of analyzed material and a broader 
range of proveniences would, of course, be desirable, it is felt that the 
results obtained by the above sampling procedures are sufficient to charac- 
terize the plant-human relationships at the Yaughan and Curriboo Plantations. 

The Nature of the Evidence 

Before discussing what plant remains were recovered from the two plantations, 
it is necessary to consider the factors involved in preserving plant remains 
in the archaeological record. Like all organic material, plant remains are 
readily devoured by a host of organisms which inhabit the soil or scavenge 
its surface. Unless special conditions deter these organisms, plant remains 
are quickly removed from the archaeological record. In certain restricted 
localities such as dry -caves, permanently wet or frigidly cold sites, envir- 
onmental conditions are sufficiently extreme as to preclude the existence of 
the destructive organisms. In such localities as these, plant remains have 
an excellent chance for preservation. Unfortunately, archaeological sites 
are infrequently located in such "protected" environments, and the typical 
open site such as the Yaughan or Curriboo Plantation can be expected to sup- 
port a full array of decay-producing organisms. Fortunately for the archaeo- 
logist, plant material which is carbonized is made Immune to such decay while 
generally retaining sufficient structure to be identifiable microscopically, 
provided the material is not mechanically destroyed. 

Because open sites preserve only carbonized plant remains, a knowledge of the 
likelihood of various plants being carbonized is of critical importance in 
interpreting the archaeological record. It should be noted that exposure to 
fire is not, by itself, sufficient to insure preservation by carbonization; 
rather, those remains which are burned in the presence of oxygen are quickly 
transformed into a fine, structureless ash. It is only those remains which 
are exposed to high heat in a reducing (oxygen deprived) atmosphere that are 
preserved with intact structure. Fortunately, the combination of high heat 
and a reducing atmosphere is found near the interior of most fires, particu- 
larly beneath the ash layer which accumulates as the fuel is consumed. It is 
from such areas as these that plant remains enter the archaeological record 
of open sites. 



G-6 



Obviously, not all plant materials utilized by a site's inhabitants are equal- 
ly likely to be preserved by carbonization. Those plant parts deliberately 
added to the fire as fuel are the most likely to be preserved. Although such 
items as corncobs and nutshells may have been used as fuel when they were 
abundant, wood is the most commonly utilized plant fuel and is almost always 
the largest component of any flotation sample. At the opposite extreme, 
plants used exclusively in areas distant from fires will usually be absent 
from the archaeological record. For this reason, certain plant foods, for 
example berries or small fruit, which may have been eaten fairly frequently 
as snacks in the localities where they grew, but infrequently transported 
back to the fireside, are probably under-represented in the archaeological 
record. 

It is probably safe to assume, however, that until recently most important 
plant foods were either prepared or consumed in proximity to a fire. Any 
inedible portion of a plant food, such as a nutshell or a peach pit, may very 
well have been disposed of in a nearby fire, and small seeds were probably 
frequently lost into the flames during food preparation. These items have an 
excellent chance of being recovered archaeologically. 

The proximity to fire in which a plant is utilized is not the only factor 
which affects the probability of a plant's being recovered archaeologically. 
Dense plant structures like seeds are more likely to be preserved in a recog- 
nizable form. Succulent plant parts such as leaves or tubers are much less 
likely to be preserved. Hence, evidence of the utilization of plants for 
greens or tubers is difficult to acquire, particularly if, like tobacco or 
indigo, the plant is intentionally harvested before it fruits, thus preclud- 
ing the fortuitous carbonization of their seeds. Weedy annuals, however, 
which thrive in areas disturbed by human activity, produce a myriad of seeds 
which are often carbonized when they are dispersed into an open fire or when 
the weedy area is intentionally burned to clear it. 

Finally, the way in which a plant is processed may greatly affect the probabi- 
lity of its being preserved. The parching of seeds over an open fire is high- 
ly likely to result in some seeds entering the fire where they may be carbon- 
ized fairly intact. On the other hand, the grinding of seeds to produce 
flour or meal may lessen the chance of any spilled portion's avoiding combus- 
tion and certainly decreases the possibility of correctly identifying the 
seeds. 

Analysis and Interpretation 

During the course of analysis, 726 grams of material were examined from 60 
provenience units. Overall the samples were rather trashy, containing a 
total of 135 grams of material other than carbonized plant remains. This was 
primarily rootlets and pieces of soil, but included an occasional uncarbon- 
ized seed or wood fragment. As the antiquity of uncarbonized remains cannot 
be demonstrated, all such remains are entered under the category "trash" and 
are not included in the other categories, which tally only carbonized materi- 
al. The carbonized material included 566 grams of wood and pitch, 2.5 grams 
of maize cupules, 0.8 gram of walnut shell, 1.0 gram of hickory nutshell, 
10.5 grams of peach pits, 1.5 grams of small seeds, and 7.8 grams of unidenti- 
fied fragments, a category which includes a mixture of rather amorphous 
pieces, some of which are probably galls, fungus or bark. 



G-7 



The majority of the carbonized plant remains recovered from the Yaughan and 
Curriboo Plantations represent plants used as fuel. This includes the wood 
and pitch^ and the maize cupules. Wood occurred in all of the features 
analyzed, and while no rigorous attempt at species identification was under- 
taken, it can be said with confidence that the overwhelming majority of wood 
fragments were pine, with hardwood fragments being extremely rare. Maize 
cupules, small cup-like structures on the cob, from whence the kernels origi- 
nate, were present in 25 of the 51 flotation samples and indicate the use of 
corncobs as fuel . 

The presence of maize cupules also strongly suggests the use of maize as a 
foodstuff, and this is further indicated by the occurrence of maize kernels 
in four of the samples. This is a surprisingly low number, considering the 
well established role of maize in prehistoric and historic period diets in 
the southern United States. Of course, the possibility that maize was not an 
important dietary item of the Curriboo and Yaughan Plantation slaves cannot 
be ruled out absolutely, but the large number of cupules in the samples and 
the regularity with which maize is mentioned as a staple food of slaves from 
mid-nineteenth century South Carolina plantations (cf Rawick 1972:14, 26, 39, 
52, 55, 62, 99, 119, and other passages) makes this, in my opinion, an im- 
plausible situation. Rather, the small number of maize kernels may be the 
result of highly effective milling which has militated against preservation. 
Rice ( Oryza sativa L.) appears to have been an important foodstuff, 19 grams 
of which occur in 12 samples. In the eighteenth century rice was an impor- 
tant cash crop in South Carolina, ranking along with indigo as one of the two 
most important sources of wealth during the period 1750 to 1776 (Wallace 
1966:188). The desirability of inundating rice fields restricted its culti- 
vation to river bottoms, and the difficulty of transporting it overland due 
to its high weight-to-volume ratio made its production near navigable water- 
ways even more attractive (Wallace 1966:189). The Curriboo and Yaughan 
Plantations would thus seem to have been well located for its cultivation. 
Rice 1s frequently mentioned as an article 1n the diet of South Carolina 
slaves (Otto 1977:103; Rawick 1972:55, 100). 

The one other cultigen which was definitely identified was the peach ( Prunus 
persica ), like rice a native of Asia. Twelve peach pits from nine samples 
were recovered, but this probably does not accurately reflect its true diet- 
ary significance. Instead of rivalling rfce or maize as a foodstuff, as its 
frequency of occurrence might suggest, peaches were likely little more than a 
dietary complement available only during a limited harvest season of June to 
July (cf Schopmeyer 1974:664). Its relative abundance in archaeological 
sites is largely due to the density of the pit, which makes it quite durable, 
and to its large size which makes it quite noticeable to excavators. 



^'Pltch" is to be understood as a generic term for any resinous substance 
exuded by wood as it burns, and does not refer to a deliberately manufactured 
naval store. 



G-8 



One other plant remain may derive from a cultigen. Feature 65 for the 
Curriboo Plantation contained a carbonized plant part, roughly discoidal with 
a diameter of 18 millimeters and a thickness of 9 millimeters. This most 
closely resembles a section of the peduncle (fruit stalk) of one of the 
Cucurbitaceae, but this identification is far from certain. 

The walnut and hickory nutshell seem to represent snack foods rather than 
dietary staples. Walnut occurred in only seven samples, and hickory in only 
ten. Furthermore, they were not a large component of any flotation sample 
but occurred in small quantities; in fact, their total combined weight of 
1.78 grams comprises only 0.3 percent of the total carbonized plant remains 
recovered. This, along with the total absence of acorns from the samples, 
seems to indicate a very limited exploitation of nuts by the site 
inhabitants. 

The hawthorn ( Crataegus sp) and bramble ( Rubus sp) seeds would also seem to 
represent dietary comp ! ements , as both were represented by only one seed 
apiece. The fruit of the hawthorn is a small pome with large seeds and a 
small amount of" pulp (Fernald and Kinsey 1958). It is therefore a food of 
limited appeal, although John Lawson in 1709 described the haws of North 
Carolina as having "... a very pleasant agreeable Taste" (Lawson 
1967:112). Bramble (a general term for the genus which includes blackber- 
ries, raspberries, and dewberries) can be quite abundant in localized areas 
and can produce a profusion of fruits during its midsummer fruiting season. 
The sole seed of Sumac ( Rhus sp) may represent a dietary item, as the seeds 
can be used to produce a pleasantly acidic beverage (Fernald and Kinsey 
1958). It is equally possible, however, that the seed may have derived from 
a nearby plant that had colonized the disturbed habitats created by human 
activities. 

The other plants which were identified from the samples are unlikely to have 
been of economic importance, but rather are weedy species which thrive in 
disturbed habitats such as those surrounding human habitations. The 13 
legume seeds seem to fit into this category rather than being domesticated 
beans. The legume from 75F2 may be rattlebox ( Crotalapia sp.), but it is too 
distorted to be confidently identified; and the five legumes from 76BF82 may 
possibly be Strophostyles sp., but the identification is far from certain. 
Two other "weedy genera , Rumex and Acalypha, are represented by one seed 
each; and four seeds are possibly Euphoriba col lata . These plants invade 
disturbed habitats, so their presence around plantation slave quarters is 
hardly remarkable. 

The Polygonum seed, 1.6 millimeters long, and trigonous with concave sides, 
is probably Polygonum hydropiperoides . This species inhabits swamp forests, 
streams, and dicthes (Radford, Ahles, and Bell 1968). So its presence in a 
canal is understandable. How it became carbonized is more problematic, but 
is an indication of fires located outside of the domestic structures. Fires 
may have been used to clear areas of weedy growth, or the seed may have been 
dispersed into a fire used for some other outdoor activity such as boiling 
laundry, making soap, or burning rubbish. 



G-9 



Several grass seeds were also found in the samples. Of special interest are 
the three carbonized seeds of goosegrass ( Eleusine indica L.) . Goosegrass is 
a common grass in the Carolinas today (Radford, Ahies and Bell 1968:116), but 
is a native of Asia (Martin 1972:19). The three seeds found in the early 
nineteenth century Yaughan Plantation samples are the earliest evidence of 
its occurrence in the New World of which I know. 

The "unidentified type one" is the most numerous grass seed found, and the 
most troubling. It is roughly cylindrical with a beveled end and a shallow 
groove along one side. Its classification as a grass seed is somewhat ques- 
tionable, resting on the interpretation of the beveled end as a basal embryo 
area. However, the bevel and the shallow groove are on the same side of the 
seed, an arrangement not found on any grass with which I am familiar. Fur- 
thermore, the seeds are highly variable in size, ranging from 2.3 millimeters 
to 4.2 millimeters long. It is possible that this seed type is not a grass, 
and may, in fact, not be a seed. 

The other "unidentified grasses" category includes one seed of either Setaria 
or Paspalum from 245K4-1. Identification cannot be more certain, since the 
seed is both distorted and eroded. The other four seeds in this category are 
fragmentary remains of small grass seeds such as Pani cum or Digitaria . Like 
the other weedy plants identified, the grasses are likely to have been coloni- 
zers of disrupted areas of the plantations, and whose seeds were most likely 
carbonized fortuitously rather than as a result of any human utilization of 
them. 

The number of unidentified seeds is quite large compared to Amerind sites. 
These seeds are typically minute and fragmentary and are, for the most part, 
"unidentifiable" as opposed to merely "unidentified". No one type (other 
than the "unidentified grass type 1") occurs with any apparent regularity or 
in significant numbers in any one feature. This suggests that the seeds are 
derived from local weeds rather than from economic plants, since seeds util- 
ized to any great extent generally occur both frequently and in concentra- 
tion. This, at least, is true of Amerind sites. 

Conclusions 

Probably the most general conclusion that can be drawn from this study is 
that pal eoethnobotan leal analysis can be fruitfully allied with historic 
sites archaeology. The archaeological record of the Yaughan and Curriboo 
Plantations was demonstrated to hold a significant amount of carbonized plant 
material which could be collected with an inexpensive and uncomplicated flo- 
tation apparatus, and the analysis of the material has added to our knowledge 
of the plant-human relationships existing on an antebellum plantation. 

It might be expected that a study of plant usages on an antebellum plantation 
might be most fruitfully studied through an analysis of archival records. 
While such records are an important source of information and should not be 
ignored, it is generally the case that archival records and the archaeologi- 
cal records complement each other. Archival records deal primarily with the 
plantations' cash crops, the plants which obviously necessitate the most 
record-keeping. For information concerning such crops as indigo, cotton or 
tobacco, archival records can be quite informative. 



G-10 



On the other hand, a paleoethnobotanical study such as this one has much less 
chance of gaining information concerning cash crops. This is hardly sur- 
prising, however, as excavations centered on domestic areas are not likely to 
encounter evidence of the processing, storage or transporting of cash crops, 
as these activities were probably carried on in areas of the plantation re- 
moved from the domestic structures. An expansion of the excavations to 
include other areas of the plantations may detect archaeological evidence for 
particular cash crops, but more likely in the form of structures associated 
with their storage or processing than in remains of the plants themselves. 
For example, the extraction of the blue pigment from indigo ( Indigofer 
tine tori a ) requires a three-tiered system of brick and mortar vats in which 
the plants are boiled, fermented, and macerated (Crokatt 1746; 1747). The 
vats are more likely to be recognizable than the plant itself, which is 
described as "looking like dung" (Crokatt 1746) after treatment is completed. 
While the extraction of pigments is perhaps the most extreme deformation to 
which any plant is subjected during processing, the production of fibers or 
oil is probably nearly as thorough in guaranteeing that a plant is rarely 
identified archaeologically. When one takes into account the rigors of 
processing, the spatially distinct areas associated with their processing and 
storage, and the small likelihood of their being carbonized except through 
rare conflagrations, it is hardly surprising that no cash crops, except 
possibly rice, were identified in this study. 

Paleoethnobotany can, however, provide much evidence about the plant foods 
used by a site's inhabitants. Archival evidence is inferior in this regard, 
since plants used for subsistence purposes generally invoke much less 
record-keeping, particularly if the plants are grown on the same plantation 
on which they are consumed, and of course, the utilization of wild plants 
would escape record-keeping entirely. 

It is in gaining information concerning the subsistence practices of the 
Yaughan and Curriboo Plantation slaves that this study has been most success- 
ful. Otto (1980:9-10) found that at Cannon's Point Plantation, St. Simons 
Island, Georgia, the slaves augmented their diet to a great extent by the 
hunting of wild animals; thus one might expect wild plants to have played an 
important role in the diet of the Yaughan and Curriboo Plantation slaves. 
This, however, does not appear to have been the case. Cultivated plants seem 
to have provided the overwhelming portion of the plant food eaten by the 
slaves, with wild plants providing only occasional dietary complements. 

Also, the range of plants utilized for food is quite small, with only seven, 
possibly eight, plants being utilized, and with only maize and rice seeming 
to be of any great importance. 

This orientation toward the exploitation of only a few plants is in sharp 
contrast to the pattern of exploitation of most Amerind sites. For example, 
at an early eighteenth century Saura village in Piedmont North Carolina, 
Wilson (1979) found evidence of the use of at least 19 food plants, with wild 
species, particularly hickory nuts, making a significant contribution to the 
diet. 



G-ll 



The reasons for the highly focal adaptation of the plantation slaves can only 
be speculated upon. It is, of course, possible that the paucity of wild 
plants is more apparent than real. The slaves may well have exploited wild 
greens such as pokeweed ( Phytolacca americana ) or goosefoot ( Chenopodiutn spp) 
which have not been preserved, but it is also quite possible that they grew 
cultivated greens such as turnip or mustard ( Brassica spp). It may have been 
the case that wild plants were of little importance due to the adequacy of 
the culti gen-derived diet which gave no motivation to gather wild foods. On 
the other hand, the lack of wild plant utilization may reflect the particular 
social status of the slaves. Effective exploitation of wild plants requires 
considerable mobility in order to visit the often dispersed locations where 
the plants occur, and considerable freedom to schedule activities so that one 
can gather the wild plants during their usually restricted harvest period. 
Slaves may not have possessed the necessary freedom of action to pursue suc- 
cessfully a subsistence strategy based on foraging. 

The overall adequacy of the slaves' diet is difficult to assess. The primacy 
of corn and rice as foodstuffs suggests a diet heavy in carbohydrates and low 
in other nutrients; but this conclusion must be tempered by the knowledge 
that other foods were probably eaten but not preserved. The most satisfac- 
tory method of assessing the adequacy of the slaves' diet would be to compare 
the plant remains from the slave quarters to those from the residence and 
kitchen of the plantation masters, with the degree to which the former 
matches the latter providing a rough measure of the adequacy of the slaves' 
diet (assuming, of course, that the masters did not suffer from chronic 
malnutrition). This comparison would be particularly useful in clarifying 
the reasons behind the absence from the slave quarter plant remains of sev- 
eral food plants — for example, watermelons, grapes, and apples — which 
should have been present on an antebellum plantation, and which should have 
been preserved if they were heavily utilized. I am tempted to speculate at 
this stage of the research that the slaves' diet was constricted, not to the 
point of chronic malnutrition, but rather to that of culinary monotony. This 
arrangement would provide the plantation owners with a healthy work force 
while preserving their monopoly over the plantations' choicest resources. 

It is with speculations rather than firm conclusions that this study must 
close. The plantation community 1s too complex and the study undertaken too 
narrow to allow more definitive results. It would be desirable to see larger 
areas of plantations excavated and more plant remains obtained and analyzed. 
Information on the diet of the other social classes inhabitating the planta- 
tions could be gathered, and perhaps information can be gleaned on plant 
usages other than dietary ones. Overall, much work remains to be done. Both 
historic sites archaeology and paleoethnobotany are young disciplines; an 
alliance between the two holds much promise for the future. 



Abbreviations Used in Tables G 1-6 



G-12 



Cellar 
Canal 
Floor 
Hearth 



K = Brick kiln 

L = Lithic scatter 

M = Pos thole /mold 

P = Pit 

T = Trench 

U = Excavation unit 

+ = Trace (less than 0.05 g) 



Table 6-1. Yaughan Plantation (38BK75): 
Plant Remains by Weight (g) 



G-13 







■5 
























9 






V 


.2 
















M 






WM 





31 




«■ 








U 


> 






.3 


*J 


V 


** 


«m 






11 


a. 








fl 


M. 


**> 


a. 


91 




31 




> 


*- fl w 






-« 


X 


3 


4> 


£ 


91 


■n 


s 


6- 


3 = - 


IM 




<*• 31 




Cm 


«£ 


EA 


M 


V 


1) 




< 3» 


-o 




— <J 


■9 


3 


W 




■^ 


V 







C -. 


'J z 




M Z 


e 


a 




> 


a. 


(fl 


c 


u 


V 


- N 




C 11 


« 




M 


w 






1) 


3 


-Z >— 


£ 


u = 




s 


3 





^ 


-■» 


> 


w 


b) s. 


0»— *» 


a 


"3 C 




. N 


e 


jf 





tm 





« 


w S > 


— fl u 


« 


— fl 


a 


•*» 


«■> 


o 


a 


•3 


u 


41 


3 O 


a e o 


u 


S '- 





■a 


a 


••. 


4) 




Ot 


Eta 


& 01 «■» 


2 < a. 


*• 


3 -. 


2 


r 


3 


s 


& 


Cfl 



ri 


C. 


1.00 


.75 


.„l 














F2 


P 


1.00 


13.93 


i 
.11 .06 


13.16 


.02 


| 


.S3 


.05 


rs 


F 


.31 


23.37 


1.32 


.49 


20.79 


.05 


.01 I 


.13 


.03 


FS 


M 


1.00 


1.20 


.44 




.76 












F7 


M 


1.00 


.34 


.13 




.21 












F9 


M 


1.00 


1.2S 


.72 ! 


.53 












F9 


N 


1.00 


3.7$ 


.92 


.09 


2.73 






• 




* 


F22 


P 


! .68 


18.30 


2.25 ! .06) 12.34 


.02 


.02 




3.S5 


.06 


F2S 


•> 


1.00 


26.43 


2.11 , 


.041 24.28 










* 


F26 


•} 


1.00 


.93 


.11 


.03 


.79 












F27 


? 


1.00 


32.17 


5.14 


.31 


22.51 


.13 


.45 


.08 


3.44 


.11 


F29 


P 


.23 


34.74 


1.92 


.21 


31.52 


.02 




.11 


.53 


.38 


F30 


P 


1.00 


23.65 


.16 


.31 


23.09 


.08 








.01 


F31 


P 


.29 


10.01 


1.30 


.IS 


3. S3 










.02 


F32 


P 


1.00 


1.05 


.26 


.02 .77 










» 


F3 3 


•* 

4 


1.00 


75.35 


27.05 


.23 47.65 


.02 




.35 




.05 


TF2 


1 


1.00 


S.68 


.06 


.02) S.S7 






' .03 


ai 


a 


1.00 


.64 


| 1 




.64 




uis i a 


1.00 I .72 .72 


1 i 
1 



a26 j 


1.00 


. 76 


.76 
















a33 





1.00 


.97 


1 


.79 






.19 


1 


U34 


a 


1.00 


1.45 




.49 




.96 




CJ42 


a 


I 

1.00 1 .11 


.11 












Q43 


a 


1.00 1 .14 


.14 


1 








TOTA 


LS 




277.73 


46.98 


2.02 


216.56 


0.34 


0.48 


0.72 


9.33 


0.79 



G-14 



1YJ.0J* 



paxjT^uapTUQ 



e^d^x^^V 



umuo6^x°d 



00 



CO 
00 

m 



c 
o 



c 



c 

A3 

en 



CM 

I 

C3 

<T3 



XSUITltf 



X 9d^L :ssisj:3 

pSTJT^UapTUQ 



ss^jt59soos 

oeums 
aXqniB^g 

tp«9d 
aoxH 

aOU9TU9AOJ<i 



o 



in 



vo 



CN 



CN 



in vo 



m 



CN 



CN 



cn 



o> 



vo 



cn in 



CN 



CN 



CN 



m cn 



in 



CN 



CN 

<n m 

m m 



o 

CO 



CN 



m 



vo 



3 3 



< 



Table G-3. Yaughan Plantation (38BK75) 
Plant Remains by Weight (g) 



G-15 







— V 


u 

s 

> 




u 

— 




n 




0) 










a 


<A > 


£ 




J 


4*J 


V 


^ 










'J 




J2 


< 




A 


-« 


<-* 


** 


u 




ta 




> 


4- W 






■•» 


5, 


3 


1) 


3 


en 


T3 




X 







IU 




■4- Jl 




2* 


£. 


Z 


u 


V 




21 




TJ 







•- > - 


•*■ 


3 


«n 




■-« 


V 




' 


V 


S 1) 


c 




- c 


C 


CJ 




> 


a. 


UJ 




■ 


u 


3 N 


jj o 




c a 


T3 




jj 


u 








•J 


3 


- > 


£ — 


^ 


u = 




U 


3 





£ 


** 




> 


ml 


w •— 


S< 4J 


■j> 


"3 ^ 


"O 


N 


3 


JC 





M 




^ 


fl 


- IQ 


— _ 


b 


— ,tj 




•«< 


— • 





•a 


fl 




u 


V 


o e 


U 


u 


c - 


3 


13 


fl 


■•* 


« 


S 




2. 


2b 


- < 


2 - 


E" 


3 i. 


2 


s 


2 


— 


— 


en 




ri 


p 


I. 00 


1.12 


. 44 


+ 


.62 




.06 


4- 




?2 


p 


.S3 


44.07 


4.90 


.40 


33.48 


.15 


.09 




.05 




?3 


p 


1.00 


2.29 


1.03 


.50 


.70 








+ 


?4 


p 


1.00 


.53 


.43 




.09 I 


i | 


.01 




rs 


p 


.51 


6.94 


.34 


.02 


6.53 












r7 


? 


1.00 


3.37 


.19 


.02 


3.53 




.03 1 




?a 


? 


.71 


48.26 


19.44 


.23 


27.73 


.05 


.10 


.10 


.49 


.02 




T9 


D 


l". 00 


10.08 


.26 


.07 


9.65 


.09 


.01 




?:o 


? 


1.00 


2.59 


.52 


.05 


2.00 


.02 


1 1 
1 


+ 




rii 


? 


1.00 


21.48 


4.66 


.19 


16.40 


.24 






+ 




F12 


? 


1.00 


15.13 


4.32 


.14 


10.27 


.43 








.02 




?13 


P 


1.00 


5.33 


2.15 


.02 


3.03 


.12 










?I4 


P 


1.00 


5.57 


2.31 


.02 


3.20 


.02 


1 


.02 




?15 


? 


1.00 


12.15 


1.07 


.13 


10.31 


.03 


.06 






.06 




F16 


T» 


1.00 


1.69 


.03 


.12 


1 
1.541 




?92 


s 


.41 


31.57 


.57 


.33 


30.26 


.16 




.02 


! .23 




?sa 


T 


1.00 


2.33 


.11 


.021 


2.13 


.07 




r:o6 


? 


1.00 


.09 




.09 




| 


TOTAL 






215.19 


42.37 


,u 


167.25 


1.33 


0.25 


0.21 


0.49 


0.42 





G-16 



i i 



TVAOi 



(A 


pstjT's.uapTun 


0) 




to 


eydAfBDY 


J-I 




'-O 


unnuoDAXO«j 


CD 
CO 




m 


xsumn 


c 




o 


ssssaao -xaqq-o 


c 

T5 


paTjT^uapxun 


Q_ 




C ■ 


ss«j£> as 009 


en 




3 


Sduia6»i 


^r 


scums 


t 

CD 




<U 


STqurejg 







ojo^abh 



qscad 



3ZTSW I 



33 TH 



SCUaTUSACUd 



i..Cu 



u 
u 



0|— N 

p»la3ler\ 

fa. I fa. I fa. I fa. I fa. Ifa. i fa. I fa. I fa. 



10 I 
u 

CO 

u 



us < 

m ^lt/ll\ol<"MlaO Ol S- 

fa.lfa.lfa.lfa.1 7- 



I i 



TABLE G-5. Curriboo Plantation (38BK245) 
Plant Remains by Weight 



G-17 







_ 




























u 




























a <z 


■« 


























y 3» 


a 


























«• -^ 


M 


























c y 


>■ 


























3 3 


MM 






3 


£ 


















tj 


fl 






^« 


U 


•fl 




*< 










y 


VJ > 


C 






J 


w 


y 


*■ 


** 








y 




j 


< 






fl 


-«* 


m» 


^ 


1) 




<n 




> 


w ^» 








•^ 


3> 


3 


a 


,£ 


a 


n 




i 







IM 






*< a 




3, 


^ 


•J5 


w 


9 




a 




9 









■— ~> 


"3 


3 


en 




•^ 


V 




•3 


C 9 




g 




~> - 


C 


u 




> 


0. 


■Ji 




-• 




M 


w 







e y 


n 




w 


U 








•j 


3 


— > 


^ 




£ 


41 = 




u 


3 





m\ 


tm 


m 


> 


U 


w — 


o»-u 


a 


-a 3< 


1 


N 


s 


JC 


y 


** 


V 


a 


■a 


- fl 


«M 


u 


« 


- -3 


«• 


«• 


y 


4 


<9 


& 


m 


S 


s 


9 





w 


e u 





fl 


fl 


•«• 


y 


-» 


w 


eu 


Sit 


a. < 


3 


0. 


e- 


3 i. 


2 


s 


2 


s 


a. 


•■a 


O 



pi 


P 


.4S 


ll.Slj .03 


.13 11.37 


.01 










n 


? 


.45 


1 
23.021 .30 


.13 22.05 


.02 


1 

* 




?9 


C 


1.00 


9.49J .32 


.11 9.76 


.05 






.13 


* 




?12 


P 


.33 


56. 0$ .36 


.sal 54.37 


.56 I .071 




.12 




r$o 


P 


1.00 


1.99 .09 


J 

.04 1.73 






.01 




.03 




F61 


? 


1.00 


1.73J .03 


.Ol] 1.67 










m 




P62 


P 


1.00 


12.33 .12 


1 
.19 11.96 


.02 




.04) 


♦ 




F63 


? 


.18 


9.05 .77 


.11 


3.13 


.02 








.02 




f«3 


P 


1.00 


U.ljj 


.21 


11.32 


.11 








.01 


♦0.54 q 
a*duncle? 


Al-2 


a 


1.00 


4.39 4.22 




.63 












*O.01 <y 

crass racftis 


U-3 


3 


1.00 


. 79< .64 


* 


.IS 


4b 










Al-4 


3 


1.00 


- 571 .05 


.Oil .52 






1 


• CT1 


T 


1.00 


17.40 


5.75 


.49 11. IS 










.05 




C?2 


r 


1.00 


17.661 10.06 


. S4l 7.05 




! .01 




C73 


, 


1.00 


3. 34J .47 


.271 2.59 


1 






.01 




C?7 


3 


.29 


22.I2! 15.57 


I I l 
.391 6.161 






4b 




S2-2 


K 


1.00 


.37* .76 


.02 


! 

.09 1 






K4-L 


X 


1.00 


17.43 


3.37 


.131 13.381 




1 * 




XF2 


K 


1.00 


10. 50l 2.14 


• Oil 9.35 




1 ! 




TOTAL 






232.341 


45.54 


3.430.31.98 


0.79 J 0.07 0. OS 0.13 1 . 2S 




GRAND 

TOTAL 






725.76 


135.39 


7.7S 


565.30 


2.51 


0.30 


0.98 


10.50 


1.46 





G-18 



! I I I 

i Mi 



TYiOX 



ps-cjT^uspTun 



BydXfBSY 



umuooAXO<£ 



xaumn 



PST JTaUBOTUn 



SSBXD3S009 

ssuirtDaq 

aiqnrejg 

140 93d 
3ZTBW 

33U3TU3AOJd 



— KNIffl 

CblbiCblCi. 



I— 131 



en - 
en o 

C .S 

a. 24 

3 



I ! 
i ! 



Ci.iU.IU.IU. 



in 

VOl- 

:cui< 



— iCi. 



I i 



CbiCulCb 
OICJIU 









§ £ 



a - 
z < 



o 

O B- 



G-19 



REFERENCES CITED 



Crokatt, James 

1746 Observations Concerning Indigo and Cochineal, Also on Silk, Rice, 
Pitch, Tar and Turpentine " London. 

1747 Further Observations Intended for Improving the Culture and 
Curing of Indigo in South Carolina"^ London. 

Fernald, Merritt L. and Alfred C. Kinsey 

1958 Edible Wild Plants of Eastern North America . Second edition, 
revised by R. C. Rollins. Harper and Row, New York . 

Lawson, John 

1967 A New Voyage to Carolina . Edited by H. T. Lefler. University of 
North Carolina, Chapel Hill . 

Martin, Alexander C. 

1972 Weeds . Golden Press, New York. 

Otto, John S. 

1977 Artifacts and status differences — a comparison of ceramics from 
planter, overseer, and slave sites on an antebellum plantation. 
In Research Strategies in Historical Archeology , edited by 
Stanley South. Academic Press, New York, pp. 91-118. 

1980 Race and class on antebellum plantations. In Archaeological 
Perspectives on Ethnicity in America: Afro-American and Asian 
American Culture History, edited by R. S. Schuyler. Baywood 
Monographs in Archaeology 1:3-13. 

Radford, Albert E., Harry E. Ahles, and C. Ritchie Bell 

1968 Manual of the Vascular Flora of the Carolinas . University of 
North Carolina, Chapel Hill. 

Rawick, George P. (editor) 

1972 The American Slave: A Composite Autobiography. Vol. 2. South 
Carolina Narratives. Parts 1 and 2. Contributions in 
Afro-American and African Studies No. 11. Greenwood Pub 1 i shi ng , 
Westport, Connecticut. 

Schopmeyer, C. S. 

1974 Seeds of Woody Plants in the United States. Uni ted States 
Department of Agriculture, Agriculture Handbook No. 450. 

Wallace, David D. 

1966 South Carolina: A Short History . University of South Carolina, 
Columbia. 

Watson, Patty Jo 

1976 In pursuit of prehistoric subsistence: a comparative account of 
some contemporary floatation techniques. Midcontinental Journal 
of Archaeology 1:77-100. 



G-20 



Wilson, Jack H. 

1979 European contact and plant food subsistence among the Carolina 
and Virginia Siouans. Paper presented at the 36th Annual Meeting 
of the Southeastern Archaeological Conference. 

Yarnell, Richard A. 

1974 Plant food and cultivation of the Salts Cavers. In Archaeology 
of the Mammoth Cave Area , edited by Patty Jo Watson" Academic 
Press, New York, pp. 113-122. 



APPENDIX H 

FAUNAL REPORT 

ELIZABETH REITZ 
AND 
KAY WOOD 



H-2 



FAUNAL REPORT FROM THE COOPER RIVER 
RED I VERS ION CANAL PROJECT, 1980 

Elizabeth J. Reitz and Kay Wood 

The materials analyzed in this report were excavated from three sites in 
Berkeley County, South Carolina. They are 30 miles inland from the Atlantic 
Ocean, due north of Charleston, and about one mile south of the Santee River. 
Excavations were done under the direction of Patrick H. Garrow and Thomas R. 
Wheaton of SSI, and funding was provided by the Charleston District of the 
Corps of Engineers. Interagency Archeological Services, Atlanta, adminis- 
tered the project for the Charleston Corps. The sites were to be impacted by 
the Cooper River Rediversion Canal Project. 

The three sites include 38BK75, 38BK76, and 38BK245. The first two sites 
were areas in a slave quarter on Yaughan Plantation, and the third was a 
slave quarter on Yaughan Plantation, and the third was a slave quarter area 
on neighboring Curriboo Plantation. Both plantations were established by 
French Huguenots in the 1740s and were occupied through the 1820s. These 
were rice and indigo plantations. Site 38BK75 was a plowed field at the time 
of excavation and 38BK76 had been logged. A swamp and small creek border 
these two sites. Most of the materials are from features, or block excava- 
tions associated with slave cabins. No fauna! materials are from wells. 
Site 38BK245 had been both plowed and scraped to subsoil prior to excavation. 
Most of the materials also were recovered from block excavations associated 
with an office or from features. All materials were recovered using 1/4-inch 
screen, in the field, and a tea strainer or 1/8-inch screen during flotation. 

The faunal materials from these sites were identified and analyzed by Kay 
Wood and Elizabeth J. Reitz using the comparative skeletal collection at the 
Zooarchaeology Laboratory, Department of Anthropology, University of Georgia, 
Athens. Standard zooarchaeological procedures were used as the bones were 
identified, counted, and weighed. The principle of paired elements as dis- 
cussed by Donald Grayson (1973) was used to determine minimum number of 
individuals (MNI). 

The results of the work were disappointing primarily because of the condition 
of the bone. The bone had been subjected to a great deal of post-deposition- 
al disturbance due to logging and other agricultural activities. At 38BK245 
the bones had been exposed by recent scraping activities and allowed to bake 
in the sun for over a year. As a result the bones had fused to the clay 
matrix. Efforts to extract the bone were impossible because of the fragile 
condition of the bone itself due to other factors. 



H-3 



The mean soil pH of the sites is as follows: 38BK75: soil -5.99, 
features-5.56; 38BK76:soil-5.96, features-6.37; 38BK245:soil-3.97, 

features-3.59 (Thomas R. Wheaton, personal communication). In recent tests 
at the zooarchaeol ogy laboratory it has been found that the natural pH level 
of mammalian bone is 7.0 to 8.0 and fish bone may be more base than mammal 
bone. This work will be pursued in the future, but it appears that in this 
case, with acidic soils and neutral bone, bone preservation would be poor, 
especially at 38BK245. As a result, identifications were difficult and in 
many cases impossible. All of the bone weights should be viewed with sus- 
picion. Additionally most evidence of butchering techniques, element dis- 
tribution, food processing methods, etc., is absent. Further, it is diffi- 
cult to assess the degree to which the faunal assemblage indicate human be- 
havior at the time of deposition or post-depositional events and differential 
preservation. 

There is very little in the three faunal collections that elicits special 
comment. At 38BK75 (Tables 1 and 2) most of the bone identifiable to species 
were from features. Several had been burned and a deer bone had been cut. 
This might have been a recent cut however. The catfish ( Ictalurus sp.) spine 
fragments indicate some utilization of fish resources. Anserinae include 
Canada goose, Brant's goose, and White-fronted goose. If these bones are 
from this subfamily they probably are from a Canada goose ( Branta 
canadensis ). Unfortunately, these birds were both wild and tamed (Johnson 
and Brown 1903), so that it is not possible to classify these as domestic or 
wild resources. Otherwise, domestic resources are the major faunal compo- 
nent. The materials from 38BK76 (Tables 3 and 4) differ from the neighboring 
site on the Yaughan Plantation in that oyster is present, and a human molar 
was identified. Species identified from 38BK245 are somewhat more diverse 
(Tables 5 and 6). Another set of catfish dorsal spines were identified, as 
was an opossum tibia, indicating some use of wild resources. Some of the 
bones were burned, and three were cut, although the cow radius may have been 
cut recently. 

Since highly acidic soils are compounded by mechanical disturbances, it seems 
reasonable to assume that the faunal patterns observed here are primarily the 
result of post-depositional actions rather than selective use of fish, birds, 
and mammals by the slaves. However, four points might be raised. First, 
since soil conditions at 38BK75 and 38BK76 were more favorable for bone pre- 
servation, yet there was actually less bone recovered from Yaughan plantation 
than at Curriboo, it would appear that there was actually less bone original- 
ly deposited in the contexts excavated at Yaughan. Secondly, documentary re- 
search for these plantations indicates that dietary supplements of meat pur- 
chased through the commercial market were very low (Thomas Wheaton, personal 
communication). The absence of pigs and cattle may be an indication of human 
behavior as well as preservation suggesting that domestic meat was not a ma- 
jor part of the diet. This is a substantially different pattern from that 
observed for nineteenth century plantations on the Georgia sea islands (Gibbs 
et al . 1980). More in keeping with the expectations formed from the sea 
island research, is the presence of armaments and fishing equipment c Some 
type of fishing or hunting equipment was found at both plantations (Thomas 
Wheaton, personal communication). These materials indicate that the slaves 
did to some extent exploit wild resources, an activity which is not well 



H-4 



documented in the faunal record. Finally, the ethnobotanical data and doc- 
umentary research indicate that plant foods were a more significant part of 
the diet than animal foods (Thomas Wheaton, personal communication). This 
is, of course, to be expected since most human populations do depend more 
upon plants than animals as the major caloric source (Moran 1979). 

The original intent of the research had been to compare these faunal materi- 
als with other faunal collections from slave cabin proveniences, such as that 
done by John Otto (1975). In all honesty it cannot be done. The disparity 
in volume among the collections is one factor. This might be discounted, 
however, if the bones from the Yaughan and Curriboo Plantations were in 
better condition. Unfortunately, due to depositional factors and bone at- 
trition, there is no confidence on the part of either author that these bones 
accurately represent the original faunal assemblages, or that analysis of 
these faunal components would reliably contribute to our understanding of the 
processes involved in slave subsistence. 

As can be seen from the species lists and the above discussion, the results 
of the identification are incomplete. Preservation and small sample size 
both contribute to an unreliable picture of species utilization and habitat 
exploitation. It is clear that some use was made of marine invertebrates, 
although these may have been exclusively building materials rather than food 
resources. Wild foods, represented by opossum, deer, and catfish were used 
at the plantation to some extent. Due to the unreliable nature of the col- 
lections it would be unwise to draw conclusions about the role wild foods 
played in the diet or to what extent specific wild habitat areas were ex- 
ploited. At the moment it appears that domestic resources, either pigs or 
cattle, were the major food source. From documentary evidence it appears the 
pork or beef might have been acquired locally. Further, it is possible that 
the Cooper plantations indicate a different pattern from that faunal material 
observed for nineteenth century coastal Georgia plantations. This difference 
may be the result of time or of environment and certainly merits further 
study. 

Slave subsistence strategies are ^ery poorly understood (Gibbs et al . 1980). 
Documents from the time period need to be confirmed, clarified, and ampli- 
fied. This can only be done using archaeological materials. It is not by 
coincidence that it is proving difficult to do this since most sites occupied 
by slaves appear to have been more or less continuously used ever since ei- 
ther as living areas or fields. The resulting disturbance to slave context 
means that good samples will be few and far between. It is therefore neces- 
sary that every opportunity be explored as far as possible on the chance that 
it may prove informative. 



H-5 



BIBLIOGRAPHY 

Gibbs, Tyson; Kathleen Cargill; Leslie Sue Lieberman; and Elizabeth Reitz 

1980 Nutrition in a Slave Population: An Anthropological Examination. 
Medical Anthropology 4(2) : 175-262. 

Grayson, Donald K. 

1973 On the methodology of fauna! analysis. American Antiquity 
38(4) :432 -438. ' 

Johnson, Willis G. and G. 0. Brown, eds. 

1903 The Poultry Book . Doubleday, New York. 

Moran, Emilio F. 

1979 Human Adaptability: An Introduction to Ecological Anthropology . 
Duxbury Press, North Scituate, Massachusetts. 

Otto, John Solomon 

1975 Status Differences and the Archaeological Record - A Comparison 
of Planter, Overseer, and Slave Sites from Cannon's Point 
Plantation (1/94-1861), St. Simons Island, Georgia . Ph.D. 
dissertation, University of Florida, university Microfi 1 ms , Ann 
Arbor. 



Table H-l. Species List, 38BK75 



H-6 



Species 



Ct. 



MNI Wt, gms 

% 



Accession No. 



3hr-ll 

Mammal 

5us scrofa 
Pig 



cf . Odocoileus virginianus 
deer 

3os taurus 
cow 

Bird 

cf. Anserinae 
Goose 

Fish 

Iccalums sp. 
catfish 

Done 

Tocal 



3 
64 



2 




3 


1 


32 




2 


1 


135 




249 


9 



6.04 
161.24 
44.4 14.1 



11.1 6.5 

22.2 16.05 

0.81 
11.1 1.97 

0.51 
11.1 0.3 

112.47 
319.99 



75B-F29-6 
75-F31-1 
75-F31-2 
75-4-0 

75B-F29-6 



75B-F29-5 
7 5-F31-6 



75-F29-5 



75B-F29-1 



H-7 



Table H-2. Burned and Cut Bone from 383K75 



3urned: 



Accession •'•' 








Ct. 


Taxon 


75B-F7 


1 


Bone 


753-F29-1 


6 


Fish 


75B-F29-L 


6 


Bone 


75B-F29-5 


18 


Bone 


75B-F29-7 


2 


Mammal 


75-F31-1 


_4 


Bone 


Tocal 


37 





Cut : 

Accession // Ct. Taxon Description 



75B-F29-6 1 cf. Odocoileus rt. astragalus, proximal end 

virginianus cut appears recent. 



H-8 



Table H-3. Species List, 38BK76 



SDecies Ct. MNI Wt, gms Accession No. 

9 % 



Shell 39 15.45 

Gastropod 3 0.6 76-L10-0 

76B-F72 
76A-U21-1 

cf. Trassostrea virginica 1 3.81 76A-F43-1 

oyster 

Mammal 79 132.69 

Homo sapiens 2 1.51 76-F8-1 

1 1 50% 2.63 76B-F87-1 

2 188.96 76B-U28-2 

76-F8-2 

2 1 50% 28.78 76B-U28-1 
cow 76B-U31-1 

3one 114 19.56 

Total 243 2 393.99 





human 


Sus scrofa 

Pig 


cf. 


Bos. taurus 




cow 


Bo 


> taurus 



H-9 



Table H-4. Burned and Cut Bone from 383K76 



Burned: 

Accession # 


cc. 


Taxon 


76B-U30-1 


2 


Bone 


7 63-U28-2 


2 


Mammal 


76-F8-1 


7 


Mammal 


76-12-2 


1 


Bone 


Total 


12 





Cue : 

Accession 



CC. 



Taxon 



Descripcion 



76A-U27-1 
7 6A-U16-2 
76B-U28-2 
Total 



Mammal possibly a cut mark on a bone fragment 

Mammal semi-oval cut on a bone fragment 

cf. Bos taurus rt. radius, looks recent 



H-10 



Soecies Ct. MNI Wt, gms Accession No, 

9 % 



Bone _53 151.16 

Total 401 10 13*7.16 



Table H-5. Species List, 38BK245 



H-ll 



Species 



Ct 



MNI 



Wt , gms 



Accession No, 



Shell 
Gastropod 



128 
9 



cf . Oassostrea 
virginica 
ovscer 



opossum 



1 



Morcenaria 
mercenaria 
Quahog clam 

Mammal 170 

cf . Didelphis 1 
vir^iniana 



cf . Sus scrof a 1 
Pig 

Sus scrof a 4 
Pig 

cf . 3os taurus 1 

cow 

Bos taurus 10 

cow 



Snake 


1 


Fish 


16 


Ictalurus sp. 


2 


catfish 





10% 



30 Z 



30% 



10% 



20% 



318.93 



0.17 



31.44 



31.22 



6.29 
21.58 

24.72 
233.93 

0.2 

0.33 

0.01 



245C-U3-2U) 
245C-U3-2(3) 
245C-U3-2(3) 
245C-F1-4 

245-F1-2 
245C-U2-K2) 



245-L16-5 



525.72 

1.64 245C-F1-4 



245-F3 
245-F12-1 

245- F3 

245B-F11-24 

245-F12-6 

245-L12-5 
245-F63-1 

245B-F11-2 

245B-F11-19 

245B-F62-2 

245C-U3-2(1) 



245-F12-6 
245-F12-9 



H-12 



Table H-6. Burned and Cut Bone from 38BK245 



_3 ur ned : 
Access iou ••■• Ct. Taxon 



245-F31-2 

245-F5 
245-F12-1 

245-F12-4 

245-F12-3 

245-F12-6 

245-F65 

Tocal 



2 


Bone 


2 


Bone 


1 


Bone 


10 


Bone 


L 


Bone 


7 


Bone 


_1 


Bone 


24 





Cut : 

Accession >/ Ct. Taxon Description 

245B-F11-2 1 Bos taurus radius, distal end, 

appears recent 

245-F12-3 2 Mammal Butcher marks-sawed 

carpal fragments 

Total 3 



Table H-7. Breakdown of Features 76B-F87 and 76B-F88 



H-13 





76B-F87: 














Accession 


# 


Shell 


.Sus 


scrofa 


Bone 


MNI 






ct. wt. 


ct. 


wt. 


ct. wt. 




F87-1 








2.63g 


1 0.36g 


1 


F87-2 




1 O.lg 






1 Tr 




F37-5 










2 O.llg 




F87-8 




3 0.06g 










F87-9 




7 0.22g 










F87-10 




3 0.09g 
14 0.47g 










Total 


2.63g 


4 0.47g 


1 


76B-F88: 














Accession 


9 


Shell 
ct. wt. 


Bone 

ct. wt. 


MNI 




F88-5 






3 


0.04g 






F88-10 




1 Tr 










F88-11 




1 0.14g 

2 0.14g 


_1 
4 


0.07g 
O.llg 






Total 









DEMCO 38-297 



UNIVERSITY OF GEORGIA LIBRARIES 



^ 



3 ElDfi DM73S 2561